ggfox00000's picture
Upload ICSI full mirror (75 wav + ICSI NXT + ICSIplus contrib + MRT)
e0cf8ff verified
<?xml version="1.0" encoding="iso-8859-1"?>
<!DOCTYPE Meeting SYSTEM "http://www.icsi.berkeley.edu/Speech/mr/Meeting1.dtd">
<Meeting Session="Bmr019" DateTimeStamp="2001-03-29-1400" Version="2" VersionDate="Jan 7 2004">
<Preamble>
<Notes>
fe016 and mn005 arrived late. (~12min, ~15min, resp.)
Digits were read simultaneously. No digits by me001 (perhaps due to
microphone problem). Digits transcript numbers were not read for two of the
speakers: L-47 for fe016 and L-42 for mn005.
Channel 8's microphone has significant microphone noise, becoming especially
bad starting at 41 minutes, 46 seconds. This is the beginning of the
microphone's death; after a stretch of bad noise, nothing more is heard
on this channel except for occasional minor noise -- no further speech,
laughter, etc. from the speaker (including no digits).
</Notes>
<Channels>
<Channel Name="chan0" Mic="c1" Transmission="w1" Gain="9" AudioFile="chan0.sph"/>
<Channel Name="chan1" Mic="s1" Transmission="w1" Gain="9" AudioFile="chan1.sph"/>
<Channel Name="chan2" Mic="s1" Transmission="w1" Gain="9" AudioFile="chan2.sph"/>
<Channel Name="chan3" Mic="s1" Transmission="w1" Gain="9" AudioFile="chan3.sph"/>
<Channel Name="chan4" Mic="s1" Transmission="w1" Gain="9" AudioFile="chan4.sph"/>
<Channel Name="chan5" Mic="c1" Transmission="w1" Gain="9" AudioFile="chan5.sph"/>
<Channel Name="chan6" Mic="c2" Transmission="u1" Gain="9" AudioFile="chan6.sph"/>
<Channel Name="chan7" Mic="c2" Transmission="u1" Gain="9" AudioFile="chan7.sph"/>
<Channel Name="chan8" Mic="u1" Transmission="j1" Gain="11" AudioFile="chan8.sph"/>
<Channel Name="chanA" Mic="a1" Transmission="j1" Gain="7" AudioFile="chanA.sph"/>
<Channel Name="chanB" Mic="s1" Transmission="j1" Gain="12" AudioFile="chanB.sph"/>
<Channel Name="chanC" Mic="u3" Transmission="j1" Gain="12" AudioFile="chanC.sph"/>
<Channel Name="chanD" Mic="u3" Transmission="j1" Gain="12" AudioFile="chanD.sph"/>
<Channel Name="chanE" Mic="c2" Transmission="u1" Gain="9" AudioFile="chanE.sph"/>
<Channel Name="chanF" Mic="c2" Transmission="u1" Gain="9" AudioFile="chanF.sph"/>
</Channels>
<Participants DB="icsi1.spk">
<Participant Name="fe016" Channel="chan0" Seat="6"/>
<Participant Name="me013" Channel="chan1" Seat="7"/>
<Participant Name="fe008" Channel="chan2" Seat="4"/>
<Participant Name="mn014" Channel="chan3" Seat="2"/>
<Participant Name="me011" Channel="chan4" Seat="9"/>
<Participant Name="mn017" Channel="chan5" Seat="5"/>
<Participant Name="me001" Channel="chan8" Seat="8"/>
<Participant Name="mn005" Channel="chanA" Seat="3"/>
<Participant Name="me026" Channel="chanB" Seat="1"/>
</Participants>
</Preamble>
<Transcript StartTime="0.0" EndTime="3586.171">
<Segment StartTime="0.323" EndTime="1.460" Participant="me011">
O_K, we're on.
</Segment>
<Segment StartTime="1.286" EndTime="1.840" Participant="me001">
<NonVocalSound Description="tapping on mike"/>
</Segment>
<Segment StartTime="1.622" EndTime="2.631" Participant="me013">
O_K. <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2.304" EndTime="5.340" Participant="me011">
So, I mean, everyone who's on the wireless check that they're on.
</Segment>
<Segment StartTime="2.354" EndTime="2.947" Participant="me001">
Alright.
</Segment>
<Segment StartTime="2.374" EndTime="2.898" Participant="mn017">
C- we -
</Segment>
<Segment StartTime="3.392" EndTime="4.123" Participant="fe008">
I see.
</Segment>
<Segment StartTime="4.233" EndTime="14.340" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="6.250" EndTime="6.719" Participant="mn017">
Yeah.
</Segment>
<Segment StartTime="6.488" EndTime="7.121" Participant="fe008">
<NonVocalSound Description="mike noise"/> Yeah.
</Segment>
<Segment StartTime="7.398" EndTime="8.071" CloseMic="false">
<NonVocalSound Description="door"/>
</Segment>
<Segment StartTime="8.110" EndTime="10.721" Participant="me011">
O_K, our agenda was quite short. <VocalSound Description="breath"/>
</Segment>
<Segment StartTime="9.692" EndTime="12.551" Participant="me013">
<VocalSound Description="inbreath"/> Oh, could you <Pause/> close the door, maybe? Yeah.
</Segment>
<Segment StartTime="11.433" EndTime="13.741" Participant="me011">
Sure. Two items, which was, uh,
</Segment>
<Segment StartTime="14.622" EndTime="23.990" Participant="me011">
digits and <Emphasis> possibly </Emphasis> stuff on - on, uh, forced alignment, which <Emphasis> Jane </Emphasis> said that Liz and Andreas had in- information on, but they didn't, so.
</Segment>
<Segment StartTime="19.426" EndTime="21.265" Participant="fe008">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="19.485" EndTime="20.375" CloseMic="false">
<NonVocalSound Description="door"/>
</Segment>
<Segment StartTime="21.685" EndTime="23.425" Participant="me013">
<VocalSound Description="inbreath"/> <Uncertain NumSyllables="4"> @@ </Uncertain> <Comment Description="uncodeable vocal sounds, trying to gain floor"/> <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="21.903" EndTime="22.564" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="23.425" EndTime="25.720" Participant="me013">
I guess the only other thing, uh, for which I -
</Segment>
<Segment StartTime="24.432" EndTime="32.697" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="25.119" EndTime="29.020" Participant="mn017">
We should do that second, because Liz might join us in time for that.
</Segment>
<Segment StartTime="27.884" EndTime="28.589" Participant="me011">
O_K.
</Segment>
<Segment StartTime="29.287" EndTime="33.858" Participant="me013">
Um. O_K, so there's digits, alignments, and, um,
</Segment>
<Segment StartTime="34.387" EndTime="35.757" Participant="me013">
I guess the other thing,
</Segment>
<Segment StartTime="35.757" EndTime="36.865" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="36.865" EndTime="39.786" Participant="me013">
which I came unprepared for, uh, <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="39.786" EndTime="44.190" Participant="me013">
is, uh, to dis- s- s- see if there's anything anybody wants to discuss about the Saturday meeting.
</Segment>
<Segment StartTime="44.087" EndTime="44.860" Participant="me011">
Right.
</Segment>
<Segment StartTime="44.190" EndTime="45.524" Participant="me013">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="45.524" EndTime="45.996" Participant="me013">
So.
</Segment>
<Segment StartTime="46.501" EndTime="48.219" Participant="me013">
Any - I mean, maybe not.
</Segment>
<Segment StartTime="47.383" EndTime="49.855" Participant="me011">
Digits and alignments. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="48.767" EndTime="49.120" Participant="me013">
<VocalSound Description="sniff"/>
</Segment>
<Segment StartTime="49.120" EndTime="50.787" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="50.691" EndTime="51.235" Participant="me011">
But -
</Segment>
<Segment StartTime="52.030" EndTime="52.663" Participant="me013">
Uh.
</Segment>
<Segment StartTime="52.075" EndTime="56.250" Participant="mn017">
Talk about aligning people's schedules. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="53.938" EndTime="54.492" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="54.028" EndTime="55.136" Participant="me013">
Yeah. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="54.148" EndTime="55.020" Participant="me011">
Yeah.
</Segment>
<Segment StartTime="54.976" EndTime="55.411" Participant="fe008">
Mm-hmm. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="54.996" EndTime="55.609" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="55.729" EndTime="56.436" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="56.250" EndTime="58.155" Participant="mn017">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="56.436" EndTime="59.140" Participant="me013">
Yeah. I mean - Right. Yeah, I mean, it was -
</Segment>
<Segment StartTime="57.650" EndTime="62.107" Participant="me011">
<VocalSound Description="mouth"/> Yeah, it's <Emphasis> forced </Emphasis> alignment of people's schedules. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="59.140" EndTime="61.408" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="59.203" EndTime="61.540" Participant="mn017">
Yeah. If we're very - <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="59.770" EndTime="62.310" Participant="mn014">
Forced align. <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="61.408" EndTime="62.604" Participant="me013">
Yeah. <VocalSound Description="breath"/>
</Segment>
<Segment StartTime="61.998" EndTime="62.455" Participant="mn017">
<Uncertain> Yeah. </Uncertain>
</Segment>
<Segment StartTime="62.604" EndTime="69.286" Participant="me013">
With - with - whatever it was, a <Emphasis> month </Emphasis> and a half or something ahead of time, the only time we could find in common - <Emphasis> roughly </Emphasis> in common, was on a Saturday.
</Segment>
<Segment StartTime="63.330" EndTime="63.884" CloseMic="false">
<NonVocalSound Description="noise"/>
</Segment>
<Segment StartTime="67.641" EndTime="67.938" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="69.093" EndTime="69.504" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="69.286" EndTime="70.863" Participant="me013">
<VocalSound Description="inbreath"/> Ugh.
</Segment>
<Segment StartTime="69.410" EndTime="71.201" Participant="me011">
Yep.
</Segment>
<Segment StartTime="70.069" EndTime="72.710" Participant="mn017">
<VocalSound Description="breath"/> It's pretty sad. Yeah.
</Segment>
<Segment StartTime="70.441" EndTime="71.094" Participant="fe008">
<VocalSound Description="mouth"/> <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="71.793" EndTime="72.445" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="73.151" EndTime="78.155" Participant="fe008">
Have - Have we thought about having a conference call to include him in more of - <VocalSound Description="laugh"/> in more of the meeting?
</Segment>
<Segment StartTime="78.155" EndTime="80.950" Participant="fe008">
I - I mean, I don't know, if we had the - if we had the telephone on the table -
</Segment>
<Segment StartTime="80.112" EndTime="89.532" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="80.838" EndTime="83.796" Participant="me013">
No. But, h- I mean, he probably has to go <Emphasis> do </Emphasis> something. Right? <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="81.280" EndTime="87.957" Participant="mn017">
No, actually I - I have to - I have to shuttle <Pause/> kids from various places to various other places.
</Segment>
<Segment StartTime="84.690" EndTime="85.516" Participant="fe008">
I see.
</Segment>
<Segment StartTime="87.590" EndTime="88.047" Participant="fe008">
O_K.
</Segment>
<Segment StartTime="87.673" EndTime="88.516" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="87.723" EndTime="88.534" Participant="me013">
<VocalSound Description="inbreath"/> Yeah.
</Segment>
<Segment StartTime="87.957" EndTime="95.020" Participant="mn017">
So. And I don't have - and I don't, um, have a cell phone so I can't be having a conference call while driving.
</Segment>
<Segment StartTime="92.012" EndTime="92.726" Participant="mn014">
A cell phone?
</Segment>
<Segment StartTime="92.866" EndTime="94.949" Participant="me013">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="94.591" EndTime="96.253" Participant="fe008">
No. <Comment Description="lengthened"/> It's not good. That's not good.
</Segment>
<Segment StartTime="94.891" EndTime="96.663" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="94.949" EndTime="96.713" Participant="me013">
R- r- right. So we have to - we -
</Segment>
<Segment StartTime="95.239" EndTime="95.990" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="95.781" EndTime="98.609" Participant="mn017">
Plus, it would make for interesting noise - background noise.
</Segment>
<Segment StartTime="97.366" EndTime="98.336" Participant="me013">
<Uncertain NumSyllables="3"> @@ </Uncertain> <Comment Description="uncodeable, vocal squeak"/>
</Segment>
<Segment StartTime="97.938" EndTime="98.927" Participant="me026">
<VocalSound Description="laugh"/> <Comment Description="very quietly"/>
</Segment>
<Segment StartTime="98.336" EndTime="107.771" Participant="me013">
<VocalSound Description="laugh"/> So we have to equip him with a - with a - <VocalSound Description="laugh"/> with a head-mounted, uh, cell phone and - <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="98.384" EndTime="98.760" Participant="me011">
Yep.
</Segment>
<Segment StartTime="98.574" EndTime="99.432" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="98.609" EndTime="100.340" Participant="mn017">
Uh - <VocalSound Description="breath"/>
</Segment>
<Segment StartTime="101.149" EndTime="101.782" Participant="fe008">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="101.588" EndTime="108.663" Participant="me011">
Ye- we- and we'd have to force you to read lots and lots of digits, so it could get real - <VocalSound Description="laugh"/> real car noise. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="101.690" EndTime="103.040" Participant="mn017">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="104.550" EndTime="106.780" Participant="mn017">
Oh, yeah. <Pause/> Oh, yeah.
</Segment>
<Segment StartTime="104.787" EndTime="108.268" Participant="me026">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="104.859" EndTime="108.114" Participant="me001">
<VocalSound Description="laugh"/> Take advantage. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="105.275" EndTime="105.893" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="107.313" EndTime="107.787" Participant="mn014">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="108.460" EndTime="109.640" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="108.891" EndTime="109.208" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="109.061" EndTime="110.149" Participant="fe008">
<VocalSound Description="breath-laugh"/> <NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="109.309" EndTime="111.432" Participant="mn014">
And with the kids in the background. Yeah.
</Segment>
<Segment StartTime="109.410" EndTime="110.209" Participant="mn017">
I'll let - I'd let -
</Segment>
<Segment StartTime="110.850" EndTime="114.402" Participant="mn017">
I let, uh, my five-year-old have a try at the digits, eh. <VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="113.790" EndTime="114.639" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="113.944" EndTime="114.444" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="115.260" EndTime="121.220" Participant="me011">
<VocalSound Description="mouth"/> So, anyway, I can talk about digits. Um, did everyone get the results or shall I go over them again?
</Segment>
<Segment StartTime="116.192" EndTime="128.548" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="116.785" EndTime="117.042" Participant="mn017">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="121.713" EndTime="127.317" Participant="me011">
I mean that it was basically - the only thing that was even slightly surprising was that the lapel did so well.
</Segment>
<Segment StartTime="127.752" EndTime="130.827" Participant="me011">
Um, and in retrospect that's not as surprising
</Segment>
<Segment StartTime="131.200" EndTime="132.410" Participant="me011">
as maybe i-
</Segment>
<Segment StartTime="132.410" EndTime="136.083" Participant="me011">
it <Emphasis> shouldn't </Emphasis> have been as surprising as I - as - as I <Emphasis> felt </Emphasis> it was. <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="136.083" EndTime="138.849" Participant="me011">
The lapel mike is a very high-quality microphone. <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="138.849" EndTime="146.054" Participant="me011">
And as Morgan pointed out, that there are actually some advantages to it in terms of breath noises and <Emphasis> clothes </Emphasis> rustling <Pause/> if no one else is <Emphasis> talking. </Emphasis>
</Segment>
<Segment StartTime="145.878" EndTime="146.571" Participant="mn017">
Exactly.
</Segment>
<Segment StartTime="146.240" EndTime="146.689" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="146.695" EndTime="147.091" Participant="me001">
Mm-hmm. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="147.039" EndTime="148.960" Participant="me011">
Um, so, uh -
</Segment>
<Segment StartTime="148.812" EndTime="149.306" Participant="me001">
It's g- it -
</Segment>
<Segment StartTime="148.960" EndTime="154.453" Participant="me013">
Well, it's - Yeah, sort of the bre- the breath noises and the mouth clicks and so forth like that, the lapel's gonna be better on. <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="149.568" EndTime="150.567" Participant="mn014">
Or the cross-talk.
</Segment>
<Segment StartTime="153.742" EndTime="154.078" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="154.453" EndTime="159.090" Participant="me013">
The lapel is typically <Emphasis> worse </Emphasis> on the - on clothes rustling, but if no one's rustling their clothes,
</Segment>
<Segment StartTime="159.090" EndTime="163.049" Participant="me011">
Right. I mean, a lot of people are just sort of leaning over and reading the digits, so it's - it's a very
</Segment>
<Segment StartTime="159.090" EndTime="161.293" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="159.140" EndTime="159.575" Participant="mn017">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="159.592" EndTime="161.295" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="161.293" EndTime="162.990" Participant="me013">
it's - it's -
</Segment>
<Segment StartTime="163.620" EndTime="165.847" Participant="me011">
different task than sort of the natural. So.
</Segment>
<Segment StartTime="163.641" EndTime="167.659" Participant="mn014">
Yeah. <Pause/> You don't move much during reading digits, I think.
</Segment>
<Segment StartTime="165.556" EndTime="166.365" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="167.284" EndTime="167.880" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="167.390" EndTime="168.170" Participant="me011">
Right.
</Segment>
<Segment StartTime="167.785" EndTime="173.502" Participant="me001">
Probably the fact that it picks up other people's speakers - other people's <Emphasis> talking </Emphasis> is an indication of that it - the fact it <Emphasis> is </Emphasis> a good microphone.
</Segment>
<Segment StartTime="170.077" EndTime="170.495" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="173.629" EndTime="174.776" Participant="me011">
Right. Right.
</Segment>
<Segment StartTime="173.781" EndTime="178.160" Participant="me013">
Right. So in the <Emphasis> digits, </Emphasis> in most - most cases, there weren't other people talking. So.
</Segment>
<Segment StartTime="174.174" EndTime="174.708" Participant="me001">
So.
</Segment>
<Segment StartTime="175.994" EndTime="176.192" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="177.363" EndTime="180.660" Participant="mn017">
D- do the lapel mikes have any directionality to them?
</Segment>
<Segment StartTime="177.537" EndTime="193.594" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="177.833" EndTime="182.560" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="180.822" EndTime="182.162" Participant="me013">
<VocalSound Description="mouth"/> There typically don't, no.
</Segment>
<Segment StartTime="182.150" EndTime="186.758" Participant="mn017">
Because I - I suppose you could <Emphasis> make </Emphasis> some that have sort of - that you have to orient towards
</Segment>
<Segment StartTime="187.134" EndTime="189.210" Participant="mn017">
your mouth, and then it would -
</Segment>
<Segment StartTime="187.782" EndTime="189.444" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="187.911" EndTime="191.990" Participant="me011">
<VocalSound Description="inbreath"/> They have a <Emphasis> little </Emphasis> bit, but they're not noise-cancelling. So, uh -
</Segment>
<Segment StartTime="191.362" EndTime="193.854" Participant="me013">
<VocalSound Description="inbreath"/> They're - they're <Emphasis> intended </Emphasis> to be omni-directional.
</Segment>
<Segment StartTime="193.729" EndTime="194.282" Participant="me011">
Right.
</Segment>
<Segment StartTime="194.810" EndTime="195.206" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="194.835" EndTime="198.061" Participant="me013">
And th- it's - and because you don't know how people are gonna put <Pronounce Pronunciation="&quot;em&quot;"> them </Pronounce> on, you know.
</Segment>
<Segment StartTime="196.462" EndTime="196.719" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="197.249" EndTime="197.695" Participant="me011">
Right.
</Segment>
<Segment StartTime="198.051" EndTime="198.427" Participant="mn014">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="198.119" EndTime="200.295" Participant="me011">
<VocalSound Description="inbreath"/> So, also, Andreas, on that one the -
</Segment>
<Segment StartTime="200.742" EndTime="203.066" Participant="me011">
the <Emphasis> back </Emphasis> part of it should be right against your head.
</Segment>
<Segment StartTime="203.553" EndTime="206.164" Participant="me011">
And that will he- keep it from flopping aro- up and down as much.
</Segment>
<Segment StartTime="206.320" EndTime="207.419" Participant="mn017">
It is against my head.
</Segment>
<Segment StartTime="206.910" EndTime="207.679" Participant="me011">
O_K.
</Segment>
<Segment StartTime="207.835" EndTime="208.273" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="208.273" EndTime="209.244" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="209.244" EndTime="212.864" Participant="me013">
Um. Yeah, we actually talked about this in the, uh, front-end meeting this morning, too.
</Segment>
<Segment StartTime="212.632" EndTime="214.310" Participant="me011">
Uh-huh. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="212.864" EndTime="214.383" Participant="me013">
Much the same thing, and - and it was -
</Segment>
<Segment StartTime="214.383" EndTime="214.867" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="214.724" EndTime="216.564" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="214.867" EndTime="219.399" Participant="me013">
uh, I mean, <Emphasis> there </Emphasis> the point of interest to the group was <Emphasis> primarily </Emphasis> that, um,
</Segment>
<Segment StartTime="219.399" EndTime="226.951" Participant="me013">
<VocalSound Description="inbreath"/> the, uh - the system that we had that was based on H_T_<Emphasis> K, </Emphasis> that's used by, you know, <Pause/> <Emphasis> all </Emphasis> the participants in Aurora,
</Segment>
<Segment StartTime="225.009" EndTime="225.990" Participant="me011">
Everybody.
</Segment>
<Segment StartTime="226.951" EndTime="227.248" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="227.248" EndTime="228.457" Participant="me013">
was so much <Emphasis> worse </Emphasis>
</Segment>
<Segment StartTime="228.457" EndTime="232.242" Participant="me013">
<VocalSound Description="inbreath"/> than the - than the S_R_<Emphasis> I. </Emphasis> And the interesting thing is that even though,
</Segment>
<Segment StartTime="232.242" EndTime="232.567" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="232.567" EndTime="236.506" Participant="me013">
yes, it's a digits task and that's a relatively small number of words and
</Segment>
<Segment StartTime="236.506" EndTime="238.305" Participant="me013">
there's a bunch of digits that you train on,
</Segment>
<Segment StartTime="238.305" EndTime="240.206" Participant="me013">
<VocalSound Description="inbreath"/> it's just <Emphasis> not </Emphasis> as <Emphasis> good </Emphasis>
</Segment>
<Segment StartTime="239.259" EndTime="241.475" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="240.736" EndTime="246.791" Participant="me013">
as having a - a l- very large amount of data and training up a - a - a nice good big <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="246.791" EndTime="251.560" Participant="me013">
H_M_M. Um, also you had the adaptation in the S_R_I system, which we didn't have in this. <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="251.560" EndTime="253.393" Participant="me013">
Um. So.
</Segment>
<Segment StartTime="252.531" EndTime="252.729" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="253.393" EndTime="254.500" Participant="me013">
Um.
</Segment>
<Segment StartTime="254.159" EndTime="255.029" Participant="mn017">
And we know -
</Segment>
<Segment StartTime="255.274" EndTime="258.230" Participant="mn017">
Di- did I send you some results without adaptation?
</Segment>
<Segment StartTime="258.063" EndTime="261.560" Participant="me011">
No. Or if you did, I didn't include them, cuz it was -
</Segment>
<Segment StartTime="258.206" EndTime="258.443" Participant="me013">
<NonVocalSound Description="noise"/>
</Segment>
<Segment StartTime="258.443" EndTime="262.910" Participant="me013">
<VocalSound Description="mouth"/> I- s- I think Stephane, uh, had seen them. So -
</Segment>
<Segment StartTime="262.810" EndTime="264.413" Participant="mn017">
Yeah, I think I did, actually.
</Segment>
<Segment StartTime="264.413" EndTime="268.388" Participant="mn017">
So there was a significant loss from not doing the adaptation.
</Segment>
<Segment StartTime="268.590" EndTime="269.658" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="269.190" EndTime="269.714" Participant="mn017">
Um.
</Segment>
<Segment StartTime="270.340" EndTime="272.520" Participant="mn017">
A - a - a couple percent or some- I mean -
</Segment>
<Segment StartTime="273.145" EndTime="275.123" Participant="mn017">
Well, I don't know it - Overall -
</Segment>
<Segment StartTime="275.820" EndTime="277.442" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="276.132" EndTime="279.633" Participant="mn017">
Uh, I - I don't remember, but there was - <NonVocalSound Description="adjusts mike"/> there was a significant,
</Segment>
<Segment StartTime="280.107" EndTime="280.820" Participant="mn017">
um,
</Segment>
<Segment StartTime="281.136" EndTime="284.860" Participant="mn017">
loss or win <Comment Description="&quot;or win&quot; spoken away from mike"/> from adaptation - with - with adaptation.
</Segment>
<Segment StartTime="282.806" EndTime="285.756" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="285.520" EndTime="287.833" Participant="mn017">
And, um, <VocalSound Description="breath"/> <VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="288.160" EndTime="291.120" Participant="mn017">
that was the phone-loop adaptation. And then there was a very small -
</Segment>
<Segment StartTime="291.908" EndTime="295.629" Participant="mn017">
like point one percent on the natives - uh, win from doing,
</Segment>
<Segment StartTime="296.130" EndTime="296.638" Participant="mn017">
um,
</Segment>
<Segment StartTime="296.895" EndTime="300.692" Participant="mn017">
you know, adaptation to <Pause/> the recognition hypotheses.
</Segment>
<Segment StartTime="301.266" EndTime="309.070" Participant="mn017">
And <Pause/> I tried both means adaptation and means and variances, and the variances added another - or subtracted another point one percent. So, <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="309.070" EndTime="313.540" Participant="mn017">
it's, um - that's the number there. Point six, I believe, is what you get with
</Segment>
<Segment StartTime="311.925" EndTime="312.400" Participant="me011">
Right.
</Segment>
<Segment StartTime="314.272" EndTime="317.000" Participant="mn017">
both, uh, means and variance adaptation.
</Segment>
<Segment StartTime="316.972" EndTime="317.592" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="317.592" EndTime="320.625" Participant="me013">
But I think one thing is that, uh, I would <Emphasis> presume </Emphasis> -
</Segment>
<Segment StartTime="320.625" EndTime="321.989" Participant="me013">
Hav- Have you ever t- <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="321.989" EndTime="326.440" Participant="me013">
Have you ever <Emphasis> tried </Emphasis> this exact same recognizer out on the actual T_I-digits test set?
</Segment>
<Segment StartTime="328.910" EndTime="331.068" Participant="mn017">
This exact same recognizer? No.
</Segment>
<Segment StartTime="330.528" EndTime="333.994" Participant="me013">
<VocalSound Description="inbreath"/> It might be interesting to do that. Cuz my - my - cuz my sense,
</Segment>
<Segment StartTime="333.994" EndTime="336.330" Participant="me013">
<VocalSound Description="inbreath"/> um - <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="334.512" EndTime="340.191" Participant="mn017">
<VocalSound Description="mouth"/> But - but, I have - I mean, people - people at S_R_I are actually working on digits.
</Segment>
<Segment StartTime="335.930" EndTime="337.960" Participant="me011">
I bet it would do even slightly better.
</Segment>
<Segment StartTime="340.191" EndTime="343.213" Participant="mn017">
I could - and they are using a system that's,
</Segment>
<Segment StartTime="343.853" EndTime="344.440" Participant="mn017">
um -
</Segment>
<Segment StartTime="344.730" EndTime="347.160" Participant="mn017">
you know, h- is actually <Emphasis> trained </Emphasis> on digits, <VocalSound Description="breath"/>
</Segment>
<Segment StartTime="347.160" EndTime="349.834" Participant="mn017">
um, but h- h- otherwise uses the same,
</Segment>
<Segment StartTime="347.276" EndTime="347.928" Participant="me013">
<VocalSound Description="inbreath"/> <Comment Description="as if to speak"/>
</Segment>
<Segment StartTime="350.260" EndTime="351.704" Participant="mn017">
you know, decoder, the same,
</Segment>
<Segment StartTime="350.266" EndTime="350.798" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="352.130" EndTime="353.615" Participant="mn017">
uh, training
</Segment>
<Segment StartTime="354.270" EndTime="359.254" Participant="mn017">
methods, and so forth, and I could ask them what they get <Pause/> on T_I-digits. <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="356.631" EndTime="357.315" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="357.315" EndTime="363.854" Participant="me013">
Yeah, bu- although I'd be - I think it'd be interesting to just take this exact actual system so that these numbers were comparable and try it out on T_I-digits.
</Segment>
<Segment StartTime="361.388" EndTime="365.307" Participant="mn017">
Mm-hmm. Well, Adam knows how to run it, so you just
</Segment>
<Segment StartTime="363.854" EndTime="364.200" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="364.057" EndTime="365.310" Participant="me011">
Yeah. No problem.
</Segment>
<Segment StartTime="364.200" EndTime="371.773" Participant="me013">
Yeah. Yeah. Yeah. Cuz our sense from the other - from the Aurora, uh, task is that - I mean, cuz we were getting sub one percent
</Segment>
<Segment StartTime="365.821" EndTime="366.880" Participant="mn017">
make a f-
</Segment>
<Segment StartTime="366.230" EndTime="368.100" Participant="me011">
And try it with T_I-digits?
</Segment>
<Segment StartTime="368.546" EndTime="369.402" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="371.773" EndTime="372.290" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="372.151" EndTime="373.109" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="372.290" EndTime="375.910" Participant="me013">
numbers on T_I-digits also with the tandem thing. So, <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="374.953" EndTime="375.750" Participant="mn017">
Mmm.
</Segment>
<Segment StartTime="375.910" EndTime="380.699" Participant="me013">
one - so there were a number of things we noted from this. One is, yeah, the S_R_I system is a lot better than the H_T_K -
</Segment>
<Segment StartTime="380.427" EndTime="380.842" Participant="mn017">
Hmm.
</Segment>
<Segment StartTime="380.699" EndTime="384.022" Participant="me013">
<VocalSound Description="inbreath"/> this, you know, very limited training H_T_K system.
</Segment>
<Segment StartTime="383.139" EndTime="384.003" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="384.022" EndTime="387.918" Participant="me013">
<VocalSound Description="inbreath"/> Uh, but the other is that, um, the digits
</Segment>
<Segment StartTime="387.918" EndTime="388.433" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="388.433" EndTime="395.603" Participant="me013">
recorded here in this room with these close mikes, i- uh, are actually a lot <Emphasis> harder </Emphasis> than the <Pause/> studio-recording T_I-digits.
</Segment>
<Segment StartTime="395.603" EndTime="403.069" Participant="me013">
<VocalSound Description="inbreath"/> I think, you know, one reason for that, uh, might be that there's still - even though it's close-talking, there still is <Emphasis> some </Emphasis> noise and <Emphasis> some </Emphasis> room acoustics. <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="395.833" EndTime="396.765" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="401.485" EndTime="402.448" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="401.941" EndTime="402.356" Participant="me011">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="403.069" EndTime="410.142" Participant="me013">
And another might be that, uh, I'd - I would <Emphasis> presume </Emphasis> that in the studio, uh, uh, situation recording read speech that if somebody
</Segment>
<Segment StartTime="410.142" EndTime="410.615" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="410.615" EndTime="416.980" Participant="me013">
did something a little funny or n- pronounced something a little funny or made a little - that they didn't include it, they made <Pronounce Pronunciation="&quot;em&quot;"> them </Pronounce> do it again.
</Segment>
<Segment StartTime="413.888" EndTime="414.877" Participant="me011">
They didn't include it.
</Segment>
<Segment StartTime="415.531" EndTime="416.382" Participant="mn017">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="415.740" EndTime="421.456" Participant="me011">
<VocalSound Description="inbreath"/> Whereas, I took out <Pause/> the ones that I <Emphasis> noticed </Emphasis> that were blatant - that were correctable.
</Segment>
<Segment StartTime="417.130" EndTime="417.837" Participant="mn017">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="420.404" EndTime="420.859" Participant="me013">
Mmm.
</Segment>
<Segment StartTime="421.472" EndTime="422.380" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="421.931" EndTime="427.192" Participant="me011">
So that, if someone just read the wrong digit, I corrected it. And then there was another one where Jose
</Segment>
<Segment StartTime="424.740" EndTime="425.400" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="425.484" EndTime="426.885" Participant="mn017">
<VocalSound Description="cough"/>
</Segment>
<Segment StartTime="427.850" EndTime="432.795" Participant="me011">
couldn't tell whether - I couldn't tell whether he was saying zero or six. <Comment Description="rising intonation"/> And I asked <Emphasis> him </Emphasis> and he couldn't tell either. <VocalSound Description="breath"/>
</Segment>
<Segment StartTime="431.789" EndTime="432.293" Participant="me026">
Hmm.
</Segment>
<Segment StartTime="432.258" EndTime="432.597" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="432.624" EndTime="433.119" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="432.737" EndTime="433.093" Participant="mn017">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="432.795" EndTime="434.080" Participant="me011">
So I just cut it out.
</Segment>
<Segment StartTime="434.646" EndTime="438.879" Participant="me011">
You know, so I just e- edited out the first, i- uh, word of the utterance.
</Segment>
<Segment StartTime="434.793" EndTime="435.774" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="434.895" EndTime="435.310" Participant="mn017">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="438.879" EndTime="442.579" Participant="me011">
Um, so there's a <Emphasis> little </Emphasis> bit of correction but it's definitely not as clean as T_I-digits.
</Segment>
<Segment StartTime="442.579" EndTime="445.528" Participant="me011">
<VocalSound Description="inbreath"/> So my expectations is T_I-digits would, especially -
</Segment>
<Segment StartTime="446.380" EndTime="451.750" Participant="me011">
I think T_I-digits is all <Pause/> American English. Right? So it would probably do even a little better still
</Segment>
<Segment StartTime="449.341" EndTime="449.969" Participant="me013">
Mm-hmm. <VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="452.640" EndTime="454.740" Participant="me011">
on the S_R_I system, but we could give it a try.
</Segment>
<Segment StartTime="453.418" EndTime="460.269" Participant="mn017">
Well. But <Pause/> <Emphasis> remember, </Emphasis> we're using a <Emphasis> telephone </Emphasis> bandwidth front-end here, uh, on this, uh -
</Segment>
<Segment StartTime="460.269" EndTime="461.941" Participant="mn017">
on this S_R_I system, so,
</Segment>
<Segment StartTime="461.941" EndTime="462.600" Participant="mn017">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="462.600" EndTime="468.631" Participant="mn017">
um, I was - I thought that maybe that's actually a good thing because it - it gets rid of some of the - uh, the noises,
</Segment>
<Segment StartTime="469.380" EndTime="470.980" Participant="mn017">
um, you know, in the -
</Segment>
<Segment StartTime="471.444" EndTime="473.409" Participant="mn017">
the - below and above the -
</Segment>
<Segment StartTime="471.997" EndTime="473.125" Participant="me013">
<VocalSound Description="breath"/> Mm-hmm.
</Segment>
<Segment StartTime="473.820" EndTime="475.350" Participant="mn017">
um, the,
</Segment>
<Segment StartTime="475.768" EndTime="477.722" Participant="mn017">
you know, speech bandwidth <Emphasis> and, </Emphasis>
</Segment>
<Segment StartTime="476.903" EndTime="477.346" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="478.270" EndTime="478.713" Participant="mn017">
um,
</Segment>
<Segment StartTime="479.210" EndTime="485.842" Participant="mn017">
I suspect that to get sort of the last bit out of these higher-quality recordings you would have to in fact, uh,
</Segment>
<Segment StartTime="485.842" EndTime="489.066" Participant="mn017">
use models that, uh, were trained on wider-band
</Segment>
<Segment StartTime="489.066" EndTime="491.860" Participant="mn017">
data. And of course we can't do that or -
</Segment>
<Segment StartTime="491.882" EndTime="493.650" Participant="me011">
Wha- what's T_I-digits? I thought t-
</Segment>
<Segment StartTime="493.575" EndTime="498.820" Participant="me013">
It's wide-band, yeah. It's - in - in fact, we looked it up and it was actually twenty kilohertz sampling.
</Segment>
<Segment StartTime="494.500" EndTime="496.135" Participant="me011">
It <Emphasis> is </Emphasis> wide-band. O_K.
</Segment>
<Segment StartTime="494.565" EndTime="495.490" Participant="mn017">
<VocalSound Description="clears throat"/>
</Segment>
<Segment StartTime="498.260" EndTime="503.330" Participant="me011">
Oh, that's right. I - I did look that up. I couldn't remember whether that was T_I-digits or one of the other digit tasks.
</Segment>
<Segment StartTime="498.877" EndTime="499.670" Participant="mn017">
Mm-hmm. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="500.397" EndTime="500.969" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="500.852" EndTime="501.129" Participant="mn017">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="501.630" EndTime="502.122" Participant="mn017">
Right.
</Segment>
<Segment StartTime="502.765" EndTime="506.680" Participant="mn017">
But - but, I would - Yeah. It's - it's easy enough to try, just run it on -
</Segment>
<Segment StartTime="506.968" EndTime="511.422" Participant="me011">
Mm-hmm. So, Morgan, you're getting a little <Emphasis> breath </Emphasis> noise. You might wanna move the mike down a little bit.
</Segment>
<Segment StartTime="507.042" EndTime="507.545" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="508.209" EndTime="509.250" Participant="me013">
<Uncertain> See w- </Uncertain>
</Segment>
<Segment StartTime="508.289" EndTime="509.133" Participant="mn017">
Now, eh, does -
</Segment>
<Segment StartTime="509.509" EndTime="511.091" Participant="mn017">
one - one issue -
</Segment>
<Segment StartTime="511.764" EndTime="514.220" Participant="mn017">
one issue with - with that is that <VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="514.220" EndTime="520.865" Participant="mn017">
um, the system has this, uh, notion of a <Emphasis> speaker </Emphasis> to - which is used in adaptation, variance norm- uh, you know,
</Segment>
<Segment StartTime="521.345" EndTime="524.227" Participant="mn017">
both in, uh, mean and variance normalization and also in
</Segment>
<Segment StartTime="521.705" EndTime="522.061" Participant="me013">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="523.805" EndTime="526.614" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="523.841" EndTime="524.505" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="524.680" EndTime="527.142" Participant="mn017">
the V_T_L <Pause/> estimation. So -
</Segment>
<Segment StartTime="527.572" EndTime="529.980" Participant="me011">
Yeah, I noticed the script that extracted it.
</Segment>
<Segment StartTime="527.795" EndTime="528.129" Participant="mn017">
Do y- ?
</Segment>
<Segment StartTime="528.129" EndTime="528.586" Participant="mn017">
Is - ?
</Segment>
<Segment StartTime="528.586" EndTime="531.712" Participant="mn017">
So does - so th- so does - does, um, <VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="531.712" EndTime="535.480" Participant="mn017">
the T_I-digits database have speakers that are known?
</Segment>
<Segment StartTime="534.147" EndTime="535.762" Participant="me011">
Yep. <Pause/> Yep.
</Segment>
<Segment StartTime="535.937" EndTime="542.088" Participant="mn017">
And is there - is there enough data or a comparable - comparable amount of data to - to what we have in our recordings here?
</Segment>
<Segment StartTime="537.660" EndTime="539.638" Participant="me011">
<VocalSound Description="mouth"/> That I don't know.
</Segment>
<Segment StartTime="539.974" EndTime="540.983" Participant="me011">
I don't know.
</Segment>
<Segment StartTime="541.750" EndTime="543.633" Participant="me011">
I don't know how many speakers there are,
</Segment>
<Segment StartTime="542.518" EndTime="543.190" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="543.352" EndTime="544.033" Participant="mn017">
O_K.
</Segment>
<Segment StartTime="543.507" EndTime="544.120" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="543.633" EndTime="545.526" Participant="me011">
and - and how many speakers per utterance.
</Segment>
<Segment StartTime="544.911" EndTime="550.351" Participant="me013">
<VocalSound Description="inbreath"/> Well, the other thing would be to do it without the adaptation and compare to these numbers without the adaptation.
</Segment>
<Segment StartTime="550.746" EndTime="551.312" Participant="me013">
That would -
</Segment>
<Segment StartTime="550.861" EndTime="555.567" Participant="mn017">
Right. Uh, but I'm not so much worried about the adaptation, actually, than - than the, um, <VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="555.567" EndTime="556.358" Participant="mn017">
um -
</Segment>
<Segment StartTime="557.113" EndTime="558.478" Participant="mn017">
the, uh,
</Segment>
<Segment StartTime="559.072" EndTime="560.397" Participant="mn017">
V_T_L estimation.
</Segment>
<Segment StartTime="560.100" EndTime="560.476" Participant="mn014">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="560.397" EndTime="566.120" Participant="mn017">
If you have only one utterance per speaker you might actually screw up on estimating the -
</Segment>
<Segment StartTime="560.609" EndTime="561.084" Participant="me011">
Right.
</Segment>
<Segment StartTime="566.120" EndTime="569.922" Participant="mn017">
the warping, uh, factor. So, um -
</Segment>
<Segment StartTime="569.548" EndTime="570.003" Participant="me011">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="570.003" EndTime="573.160" Participant="me011">
I <Emphasis> strongly </Emphasis> suspect that they have more speakers than we do.
</Segment>
<Segment StartTime="573.410" EndTime="577.158" Participant="mn017">
Right. But it's not the amount of speakers, it's the num- it's the amount of data <Emphasis> per </Emphasis> speaker.
</Segment>
<Segment StartTime="573.623" EndTime="574.750" Participant="me011">
So, uh -
</Segment>
<Segment StartTime="576.886" EndTime="581.071" Participant="me011">
Right. So we - we could probably do an extraction that was roughly equivalent.
</Segment>
<Segment StartTime="578.150" EndTime="578.651" Participant="mn017">
Right.
</Segment>
<Segment StartTime="579.936" EndTime="580.430" Participant="mn017">
Right.
</Segment>
<Segment StartTime="581.620" EndTime="582.383" Participant="mn017">
So -
</Segment>
<Segment StartTime="581.620" EndTime="582.680" Participant="me011">
Um.
</Segment>
<Segment StartTime="583.493" EndTime="590.220" Participant="me011">
So, although I - I sort of know how to run it, there are a little - a f- few details here and there that I'll have to <Pause/> dig out.
</Segment>
<Segment StartTime="586.714" EndTime="587.039" Participant="mn017">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="587.039" EndTime="588.162" Participant="mn017">
O_K. The key -
</Segment>
<Segment StartTime="588.670" EndTime="590.591" Participant="mn017">
So th- the system actually
</Segment>
<Segment StartTime="591.000" EndTime="593.578" Participant="mn017">
extracts the speaker I_D from the waveform names.
</Segment>
<Segment StartTime="593.599" EndTime="594.568" Participant="me011">
Right. I saw that.
</Segment>
<Segment StartTime="593.993" EndTime="600.351" Participant="mn017">
And there's a - there's a script - and that is actually <Emphasis> all </Emphasis> in one script. So there's this one script that parses waveform names
</Segment>
<Segment StartTime="599.894" EndTime="603.237" Participant="me013">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="600.351" EndTime="604.052" Participant="mn017">
and extracts things like the, um, speaker,
</Segment>
<Segment StartTime="604.447" EndTime="608.067" Participant="mn017">
uh, I_D or something that can stand in as a speaker I_D. So,
</Segment>
<Segment StartTime="607.474" EndTime="611.311" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="608.067" EndTime="612.964" Participant="mn017">
we might have to modify that script to recognize the, um, speakers,
</Segment>
<Segment StartTime="611.320" EndTime="612.110" Participant="me011">
Right.
</Segment>
<Segment StartTime="612.964" EndTime="619.175" Participant="mn017">
<VocalSound Description="inbreath"/> um, in the - in the, uh, um, <VocalSound Description="mouth"/> T_I-digits <Pause/> database.
</Segment>
<Segment StartTime="613.151" EndTime="614.793" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="617.696" EndTime="618.389" Participant="me011">
Right.
</Segment>
<Segment StartTime="619.417" EndTime="620.730" Participant="me011">
And that, uh -
</Segment>
<Segment StartTime="620.399" EndTime="622.852" Participant="mn017">
<VocalSound Description="mouth"/> Or you can fake - you can
</Segment>
<Segment StartTime="623.440" EndTime="629.766" Participant="mn017">
fake <Pause/> names for these waveforms that resemble the names that we use here for the - for the meetings.
</Segment>
<Segment StartTime="628.239" EndTime="630.019" Participant="me011">
Right. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="629.766" EndTime="633.060" Participant="mn017">
That would be the, sort of - probably the safest way to do -
</Segment>
<Segment StartTime="631.530" EndTime="638.216" Participant="me011">
I might have to do that anyway to - to do - because we may have to do an extract to get the <Pause/> amount of data per speaker about right.
</Segment>
<Segment StartTime="638.147" EndTime="638.758" Participant="mn017">
Uh-huh.
</Segment>
<Segment StartTime="638.972" EndTime="642.312" Participant="me011">
The other thing is, isn't T_I-digits <Emphasis> isolated </Emphasis> digits?
</Segment>
<Segment StartTime="639.263" EndTime="639.700" Participant="mn017">
Right.
</Segment>
<Segment StartTime="642.312" EndTime="647.732" Participant="me011">
Or is that another one? I'm - I looked through a <Emphasis> bunch </Emphasis> of the digits t- corp- corpora, and now they're all blurring.
</Segment>
<Segment StartTime="648.161" EndTime="648.846" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="648.290" EndTime="652.190" Participant="me011">
Cuz one of them was literally people reading a single digit.
</Segment>
<Segment StartTime="652.469" EndTime="653.260" Participant="mn017">
<VocalSound Description="clears throat"/>
</Segment>
<Segment StartTime="653.711" EndTime="655.980" Participant="me011">
And then others were connected digits.
</Segment>
<Segment StartTime="655.808" EndTime="661.154" Participant="me013">
Yeah. Most of T_I-digits is connected digits, I think. The - I mean, we had a Bellcore corpus that we were using. It was -
</Segment>
<Segment StartTime="657.615" EndTime="658.184" Participant="me011">
O_K.
</Segment>
<Segment StartTime="660.594" EndTime="662.100" Participant="me011">
Maybe it's the Bell <Uncertain> Gram </Uncertain>.
</Segment>
<Segment StartTime="661.154" EndTime="663.508" Participant="me013">
<VocalSound Description="inbreath"/> that's - that was isolated digits.
</Segment>
<Segment StartTime="662.829" EndTime="663.621" Participant="me011">
Bell Digits.
</Segment>
<Segment StartTime="664.076" EndTime="664.590" Participant="me011">
Alright.
</Segment>
<Segment StartTime="664.633" EndTime="665.497" Participant="me013">
Um.
</Segment>
<Segment StartTime="664.770" EndTime="668.197" Participant="mn017">
By the way, I think we can improve these numbers if we care to compr- improve them
</Segment>
<Segment StartTime="668.197" EndTime="676.856" Participant="mn017">
<VocalSound Description="inbreath"/> by, um, <VocalSound Description="mouth"/> not starting with the Switchboard models but by taking the Switchboard models and doing supervised adaptation
</Segment>
<Segment StartTime="677.390" EndTime="680.381" Participant="mn017">
on a small amount of digit data collected
</Segment>
<Segment StartTime="680.667" EndTime="680.975" Participant="me011">
Yep.
</Segment>
<Segment StartTime="680.933" EndTime="681.655" Participant="mn017">
in this setting.
</Segment>
<Segment StartTime="682.089" EndTime="689.360" Participant="mn017">
Because that would adapt your models to the room acoustics and f- for the far-field microphones, you know, to the noise. <VocalSound Description="breath"/>
</Segment>
<Segment StartTime="689.360" EndTime="693.420" Participant="mn017">
And <Emphasis> that </Emphasis> should really improve things, um, further.
</Segment>
<Segment StartTime="694.136" EndTime="699.891" Participant="mn017">
And then you use those <Emphasis> adapted </Emphasis> models, which are not <Emphasis> speaker </Emphasis> adapted but sort of acous- you know, channel adapted -
</Segment>
<Segment StartTime="698.490" EndTime="699.975" Participant="me011">
Channel adapted.
</Segment>
<Segment StartTime="699.891" EndTime="703.135" Participant="mn017">
use <Emphasis> that </Emphasis> as the starting models for your speaker adaptation.
</Segment>
<Segment StartTime="703.253" EndTime="703.800" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="703.800" EndTime="706.552" Participant="me013">
<VocalSound Description="inbreath"/> But the thing is, uh - I mean, w- when you -
</Segment>
<Segment StartTime="705.393" EndTime="705.630" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="706.552" EndTime="708.708" Participant="me013">
it depends whether you're ju- were just using this as a -
</Segment>
<Segment StartTime="708.708" EndTime="716.523" Participant="me013">
<VocalSound Description="inbreath"/> a starter task for - you know, to get things going for <Emphasis> conversational </Emphasis> or if we're really interested i- in connected digits. And I - I think the answer is <Emphasis> both. </Emphasis>
</Segment>
<Segment StartTime="716.190" EndTime="718.340" Participant="mn017">
Well, I don't know.
</Segment>
<Segment StartTime="716.523" EndTime="722.142" Participant="me013">
<VocalSound Description="inbreath"/> And for - for connected digits over the telephone you don't actually want to put a whole lot of effort into adaptation because
</Segment>
<Segment StartTime="722.142" EndTime="725.465" Participant="me013">
<VocalSound Description="inbreath"/> somebody <Pause/> gets on the phone and says a number and then you just want it.
</Segment>
<Segment StartTime="722.548" EndTime="723.697" CloseMic="false">
<NonVocalSound Description="door"/> <Comment Description="[fe016] arrives"/>
</Segment>
<Segment StartTime="725.465" EndTime="726.219" Participant="me013">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="726.010" EndTime="728.370" Participant="fe008">
This is - this - that one's better. Mm-hmm. <Comment Description="whispered"/> <Comment Description="warning [fe016] away from headphones known to be faulty"/>
</Segment>
<Segment StartTime="726.219" EndTime="727.760" Participant="me013">
You don't - don't, uh -
</Segment>
<Segment StartTime="726.424" EndTime="726.820" Participant="mn017">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="726.820" EndTime="733.282" Participant="mn017">
Right. Um, but, you know, I - uh, my impression was that you were actually interested in the far-field microphone,
</Segment>
<Segment StartTime="733.915" EndTime="736.406" Participant="mn017">
uh, problem, I mean. So,
</Segment>
<Segment StartTime="736.761" EndTime="737.196" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="736.850" EndTime="741.212" Participant="mn017">
you want to - you want to - That's the obvious thing to <Emphasis> try. </Emphasis> Right? Then, eh - because
</Segment>
<Segment StartTime="738.708" EndTime="741.440" Participant="fe016">
<NonVocalSound Description="mike noise, as she puts on headphone"/>
</Segment>
<Segment StartTime="738.871" EndTime="739.326" Participant="mn014">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="739.348" EndTime="740.633" Participant="fe008">
Oh. Oh. <Comment Description="whispered, regarding [fe016]'s microphone adjustment"/>
</Segment>
<Segment StartTime="740.274" EndTime="740.803" Participant="me013">
Right.
</Segment>
<Segment StartTime="741.745" EndTime="751.172" Participant="mn017">
you - you don't have any - That's where the most m- acoustic mismatch is between the currently used models and the - the r- the set up here. So.
</Segment>
<Segment StartTime="741.838" EndTime="743.095" Participant="fe008">
Yeah. <Comment Description="whispered"/>
</Segment>
<Segment StartTime="747.589" EndTime="750.121" Participant="fe016">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="748.985" EndTime="749.420" Participant="me013">
Right.
</Segment>
<Segment StartTime="751.833" EndTime="752.328" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="752.328" EndTime="757.858" Participant="me013">
Yeah. So that'd be anoth- another interesting data point. I mean, I - I guess I'm saying I don't know if we'd want to do that as the - as -
</Segment>
<Segment StartTime="752.673" EndTime="754.233" Participant="mn017">
<VocalSound Description="outbreath"/> Mm-hmm.
</Segment>
<Segment StartTime="754.752" EndTime="755.977" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="756.295" EndTime="757.581" Participant="me011">
Other way. Liz -
</Segment>
<Segment StartTime="756.695" EndTime="757.407" Participant="mn014">
Other way.
</Segment>
<Segment StartTime="758.679" EndTime="761.725" Participant="fe016">
Now you're all watching me. <Comment Description="while laughing, not spoken into mike"/> Alright. This way.
</Segment>
<Segment StartTime="759.538" EndTime="762.340" Participant="me011">
It f- it clips over your ears. There you go.
</Segment>
<Segment StartTime="759.837" EndTime="760.592" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="760.138" EndTime="761.008" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="761.725" EndTime="763.670" Participant="fe016">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="763.144" EndTime="770.173" Participant="fe008">
If you have a <Emphasis> strong </Emphasis> fe- if you have a <Emphasis> strong </Emphasis> preference, you <Emphasis> could </Emphasis> use this. It's just we - we think it has some spikes. So, uh, we - we didn't use that one.
</Segment>
<Segment StartTime="763.670" EndTime="765.890" Participant="fe016">
You're all watching. <VocalSound Description="laugh"/> This is terrible. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="763.792" EndTime="765.453" Participant="me026">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="763.920" EndTime="766.347" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="764.123" EndTime="765.606" Participant="mn005">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="764.390" EndTime="765.083" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="764.587" EndTime="765.101" Participant="mn017">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="765.890" EndTime="766.967" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="768.802" EndTime="771.450" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="769.463" EndTime="770.774" Participant="fe016">
I'll get it.
</Segment>
<Segment StartTime="770.173" EndTime="771.340" Participant="fe008">
But you <Emphasis> could </Emphasis> if you want. I don't know.
</Segment>
<Segment StartTime="771.340" EndTime="775.123" Participant="fe008">
And Andre- Andreas, your - your microphone's a little bit low.
</Segment>
<Segment StartTime="771.450" EndTime="773.318" Participant="me013">
Yeah. At any rate, I don't know if w-
</Segment>
<Segment StartTime="772.547" EndTime="774.287" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="774.652" EndTime="776.362" Participant="mn017">
It is? Uh.
</Segment>
<Segment StartTime="774.740" EndTime="777.589" Participant="me013">
Yeah. <VocalSound Description="inbreath"/> I don't know if <Uncertain> we </Uncertain> wanna use that as the -
</Segment>
<Segment StartTime="775.123" EndTime="775.778" Participant="fe008">
Yeah.
</Segment>
<Segment StartTime="776.040" EndTime="777.876" Participant="me011">
Uh, it <Emphasis> pivots. </Emphasis>
</Segment>
<Segment StartTime="777.678" EndTime="780.823" Participant="fe008">
<VocalSound Description="inbreath"/> So if you see the picture <Pause/> and then you have to scr-
</Segment>
<Segment StartTime="778.351" EndTime="780.032" Participant="me011">
It - it - like this.
</Segment>
<Segment StartTime="778.360" EndTime="779.039" Participant="mn017">
I- I -
</Segment>
<Segment StartTime="778.595" EndTime="779.920" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="779.250" EndTime="784.552" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="779.544" EndTime="780.256" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="781.010" EndTime="781.544" Participant="mn017">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="781.544" EndTime="784.269" Participant="mn017">
I - I already adjusted this a number of times. I - I
</Segment>
<Segment StartTime="781.641" EndTime="782.096" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="783.660" EndTime="787.420" Participant="me011">
Eh. Yeah, I think these mikes are not working as well as I would like.
</Segment>
<Segment StartTime="784.680" EndTime="786.180" Participant="mn017">
can't quite seem to -
</Segment>
<Segment StartTime="787.391" EndTime="791.187" Participant="mn017">
Yeah, I think this contraption around your head is not <Pause/> working so well.
</Segment>
<Segment StartTime="788.029" EndTime="790.760" Participant="me013">
Too many adju- too many adjustments. Yeah.
</Segment>
<Segment StartTime="791.176" EndTime="791.611" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="791.611" EndTime="796.730" Participant="me013">
Anyway, what I was saying is that I - I think I probably wouldn't want to see that as sort of like the <Emphasis> norm, </Emphasis> that we compared all things to. <Comment Description="rising intonation"/>
</Segment>
<Segment StartTime="791.734" EndTime="792.836" Participant="fe008">
That looks good. Yeah. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="796.730" EndTime="800.403" Participant="me013">
To, uh, the - to have - have all this ad- all this, uh, adaptation.
</Segment>
<Segment StartTime="800.403" EndTime="802.356" Participant="me013">
<VocalSound Description="inbreath"/> But I think it's an important data point,
</Segment>
<Segment StartTime="802.688" EndTime="803.697" Participant="me013">
if you're - if - Yeah.
</Segment>
<Segment StartTime="802.745" EndTime="803.195" Participant="mn017">
Right.
</Segment>
<Segment StartTime="804.150" EndTime="805.288" Participant="me013">
<VocalSound Description="inbreath"/> Um.
</Segment>
<Segment StartTime="805.990" EndTime="809.107" Participant="me013">
<VocalSound Description="mouth"/> The other thing that - that, uh - of course,
</Segment>
<Segment StartTime="809.107" EndTime="813.294" Participant="me013">
what Barry was looking at was - was just that, the near versus far. And, yeah, the adaptation would get
</Segment>
<Segment StartTime="812.038" EndTime="812.882" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="813.294" EndTime="813.716" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="813.716" EndTime="821.135" Participant="me013">
th- some of that. But, I think even - even if there was, uh, only a factor of two or something, like I was saying in the email, I think that's - <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="821.135" EndTime="823.509" Participant="me013">
that's a <Emphasis> big </Emphasis> factor. So - <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="822.261" EndTime="822.996" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="822.794" EndTime="823.585" Participant="me011">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="823.509" EndTime="824.359" Participant="me013">
N-
</Segment>
<Segment StartTime="823.585" EndTime="827.030" Participant="me011">
Liz, you could also just use the other mike if you're having problems with that one.
</Segment>
<Segment StartTime="825.418" EndTime="826.031" Participant="fe008">
Well.
</Segment>
<Segment StartTime="826.026" EndTime="826.763" Participant="fe016">
O_K. <Comment Description="whispered"/>
</Segment>
<Segment StartTime="826.466" EndTime="831.589" Participant="fe008">
Yeah. This would be O_K. We - we - we think that this has <Emphasis> spikes </Emphasis> on it, so it's not as good acoustically, but -
</Segment>
<Segment StartTime="826.672" EndTime="830.686" Participant="mn005">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="828.208" EndTime="832.759" Participant="fe016">
It's this thing's - This is too big for my head. <Comment Description="quietly amid mike noise"/>
</Segment>
<Segment StartTime="830.356" EndTime="833.152" Participant="mn017">
Yeah, basically your <Emphasis> ears </Emphasis> are too big. I mean, mine are too.
</Segment>
<Segment StartTime="832.759" EndTime="838.053" Participant="fe016">
No, my - my - But this is too big for my <Emphasis> head. </Emphasis> So, I mean, <Comment Description="partially offmike, as she adjusts it"/>
</Segment>
<Segment StartTime="833.152" EndTime="835.459" Participant="mn017">
E- th- everybody's ears are too big for these things.
</Segment>
<Segment StartTime="836.322" EndTime="836.836" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="836.606" EndTime="837.298" Participant="mn017">
Uh -
</Segment>
<Segment StartTime="838.053" EndTime="840.470" Participant="fe016">
<Comment Description="suddenly louder, mike now adjusted"/> it doesn't - you know, it's sit- <VocalSound Description="laugh"/> <Comment Description="mike noise throughout"/>
</Segment>
<Segment StartTime="840.680" EndTime="841.099" Participant="fe008">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="841.534" EndTime="843.589" Participant="fe008">
Well, if you'd <Emphasis> rather </Emphasis> have this one then it's -
</Segment>
<Segment StartTime="841.750" EndTime="842.622" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="842.432" EndTime="845.410" Participant="fe016">
<VocalSound Description="laugh"/> O_K.
</Segment>
<Segment StartTime="843.700" EndTime="844.352" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="845.210" EndTime="846.023" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="846.509" EndTime="847.202" Participant="fe016">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="847.105" EndTime="847.600" Participant="mn017">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="847.202" EndTime="850.442" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="847.270" EndTime="847.644" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="847.880" EndTime="848.665" Participant="me011">
Oh, well.
</Segment>
<Segment StartTime="848.000" EndTime="850.295" Participant="me013">
<VocalSound Description="laugh"/> It's <Pause/> great.
</Segment>
<Segment StartTime="849.384" EndTime="850.041" Participant="me011">
So the -
</Segment>
<Segment StartTime="850.580" EndTime="854.591" Participant="me011">
To get that, uh, pivoted this way, it pivots like this.
</Segment>
<Segment StartTime="851.787" EndTime="857.596" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="854.892" EndTime="856.792" Participant="fe016">
No this way. Yeah.
</Segment>
<Segment StartTime="855.619" EndTime="856.865" Participant="me011">
Yeah. There you go.
</Segment>
<Segment StartTime="856.680" EndTime="858.104" Participant="fe008">
And there's a screw that you can tighten.
</Segment>
<Segment StartTime="857.221" EndTime="858.586" Participant="me011">
And then it - Right.
</Segment>
<Segment StartTime="858.199" EndTime="860.354" Participant="fe016">
Right. I already <Pause/> tried to get it close.
</Segment>
<Segment StartTime="859.672" EndTime="860.226" Participant="fe008">
Good.
</Segment>
<Segment StartTime="860.418" EndTime="863.210" Participant="me011">
So if it doesn't bounce around too much, that's actually good placement.
</Segment>
<Segment StartTime="862.652" EndTime="863.442" Participant="mn017">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="863.053" EndTime="866.950" Participant="fe016">
O_K. <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="863.210" EndTime="863.968" Participant="me011">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="863.212" EndTime="863.826" Participant="fe008">
That looks good.
</Segment>
<Segment StartTime="863.968" EndTime="866.381" Participant="me011">
But it looks like it's gonna bounce a lot.
</Segment>
<Segment StartTime="864.107" EndTime="868.303" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="868.303" EndTime="871.290" Participant="me013">
So, where were we? Uh - <VocalSound Description="laugh"/> Yeah.
</Segment>
<Segment StartTime="869.344" EndTime="870.560" Participant="fe008">
Yeah. <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="869.677" EndTime="872.150" Participant="me011">
Digits. Adaptation.
</Segment>
<Segment StartTime="871.290" EndTime="874.101" Participant="me013">
Uh, adaptation, non-adaptation,
</Segment>
<Segment StartTime="874.101" EndTime="874.635" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="874.635" EndTime="876.603" Participant="me013">
um, factor of two,
</Segment>
<Segment StartTime="878.057" EndTime="878.907" Participant="me013">
um -
</Segment>
<Segment StartTime="878.791" EndTime="881.550" Participant="mn017">
What k- u- By the way, wh- <Emphasis> what </Emphasis> factor of two did you - ? I mean -
</Segment>
<Segment StartTime="878.907" EndTime="880.227" Participant="me013">
Oh, yeah. I know what I was go- w-
</Segment>
<Segment StartTime="880.936" EndTime="888.788" Participant="me013">
Oh, no, no. It's tha- that - that we were saying, you know, well is - how much worse is far than near, you know. And I mean it depends on which one you're looking at, but for the everybody, it's
</Segment>
<Segment StartTime="882.156" EndTime="883.006" Participant="mn017">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="884.450" EndTime="887.261" Participant="mn017">
Oh, th- O_K. <Emphasis> That </Emphasis> factor of two.
</Segment>
<Segment StartTime="887.261" EndTime="888.110" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="888.788" EndTime="889.297" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="889.297" EndTime="890.782" Participant="me013">
little under a factor or two. <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="890.782" EndTime="893.783" Participant="me013">
Yeah. I - I know what I was thinking was that maybe, uh,
</Segment>
<Segment StartTime="893.783" EndTime="894.461" Participant="me013">
i- i-
</Segment>
<Segment StartTime="894.461" EndTime="895.272" Participant="me013">
we
</Segment>
<Segment StartTime="896.079" EndTime="904.552" Participant="me013">
<Emphasis> could </Emphasis> actually t- t- <Emphasis> try </Emphasis> at least <Emphasis> looking </Emphasis> at, uh, some of the - the large vocabulary speech from a far microphone, at least from the <Emphasis> good </Emphasis> one.
</Segment>
<Segment StartTime="903.572" EndTime="904.516" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="904.552" EndTime="905.079" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="905.079" EndTime="909.540" Participant="me013">
I mean, <Emphasis> before </Emphasis> I thought we'd get, you know, a hundred and fifty percent error or something, but if - <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="907.897" EndTime="909.561" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="909.540" EndTime="911.199" Participant="me013">
if, uh - if we're getting
</Segment>
<Segment StartTime="911.555" EndTime="913.630" Participant="me013">
thirty-five, forty percent or something,
</Segment>
<Segment StartTime="913.630" EndTime="914.032" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="913.960" EndTime="915.346" Participant="fe016">
Actually if you run,
</Segment>
<Segment StartTime="914.032" EndTime="915.239" Participant="me013">
u- um -
</Segment>
<Segment StartTime="915.999" EndTime="923.012" Participant="fe016">
though, on a close-talking mike over the whole meeting, during all those silences, you get, like, four hundred percent word error.
</Segment>
<Segment StartTime="920.494" EndTime="921.150" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="922.181" EndTime="925.805" Participant="me013">
<VocalSound Description="mouth"/> Right. I understand. But doing the same kind of limited thing -
</Segment>
<Segment StartTime="924.979" EndTime="927.500" Participant="fe016">
Or - or some high number. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="926.636" EndTime="930.501" Participant="me013">
Yeah, sure. Get all these insertions. But I'm saying if you do the same kind of limited thing
</Segment>
<Segment StartTime="928.890" EndTime="929.588" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="929.928" EndTime="943.261" Participant="mn005">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="930.501" EndTime="935.230" Participant="me013">
<VocalSound Description="inbreath"/> as people have done in Switchboard evaluations or as - a-
</Segment>
<Segment StartTime="934.263" EndTime="936.933" Participant="fe016">
Where you know who the speaker is and there's no overlap?
</Segment>
<Segment StartTime="936.941" EndTime="938.068" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="937.140" EndTime="939.810" Participant="fe016">
And you do just the far-field for those regions?
</Segment>
<Segment StartTime="940.046" EndTime="942.618" Participant="me013">
Yeah. The same sort of numbers that we got those graphs from. Right?
</Segment>
<Segment StartTime="941.640" EndTime="945.616" Participant="me011">
Could we do <Emphasis> exactly </Emphasis> the same thing that we're doing now, but do it with a far-field mike?
</Segment>
<Segment StartTime="945.367" EndTime="946.554" Participant="me013">
Yeah, do it with one of - on-
</Segment>
<Segment StartTime="945.616" EndTime="950.180" Participant="me011">
Cuz we extract the <Emphasis> times </Emphasis> from the near-field mike, but you use the <Emphasis> acoustics </Emphasis> from the far-field mike.
</Segment>
<Segment StartTime="949.493" EndTime="954.632" Participant="fe016">
Right. I understand that. I just meant that - so you have <Pause/> three choices. There's, um -
</Segment>
<Segment StartTime="955.801" EndTime="957.581" Participant="fe016">
You can use times
</Segment>
<Segment StartTime="958.234" EndTime="963.497" Participant="fe016">
where that person is talking <Emphasis> only </Emphasis> from the transcripts but the segmentations were -
</Segment>
<Segment StartTime="963.927" EndTime="967.745" Participant="fe016">
were synchronized. Or you can do a forced alignment on the <Emphasis> close-talking </Emphasis>
</Segment>
<Segment StartTime="965.334" EndTime="970.180" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="968.079" EndTime="973.494" Participant="fe016">
to determine that, the- you know, within this segment, these really were the times that this person was talking and
</Segment>
<Segment StartTime="973.494" EndTime="976.898" Participant="fe016">
<Emphasis> elsewhere </Emphasis> in the segment other people are <Emphasis> overlapping </Emphasis> and just front-end <Emphasis> those </Emphasis> pieces.
</Segment>
<Segment StartTime="977.538" EndTime="979.694" Participant="fe016">
Or you can run it on the whole data, which is -
</Segment>
<Segment StartTime="978.818" EndTime="979.451" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="980.202" EndTime="983.187" Participant="me013">
But - but - but how did we get the - how did we determine
</Segment>
<Segment StartTime="980.287" EndTime="981.362" Participant="fe016">
which is, you know, a -
</Segment>
<Segment StartTime="983.743" EndTime="984.772" Participant="me013">
the links,
</Segment>
<Segment StartTime="985.365" EndTime="987.165" Participant="me013">
uh, that we're testing on
</Segment>
<Segment StartTime="987.640" EndTime="989.123" Participant="me013">
in the stuff we reported?
</Segment>
<Segment StartTime="989.300" EndTime="997.880" Participant="fe016">
In the H_L_<Emphasis> T </Emphasis> paper we took <Pause/> segments that are channel - <Emphasis> time-aligned, </Emphasis> which is now h- being changed in the <Emphasis> transcription </Emphasis> process, which is good,
</Segment>
<Segment StartTime="997.880" EndTime="998.256" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="998.256" EndTime="1004.733" Participant="fe016">
and we took cases where the <Emphasis> transcribers </Emphasis> said there was only one person talking here, because no one else had time -
</Segment>
<Segment StartTime="998.673" EndTime="1001.978" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1004.733" EndTime="1009.007" Participant="fe016">
<Emphasis> any </Emphasis> words in that segment and called that "non-overlap".
</Segment>
<Segment StartTime="1007.385" EndTime="1007.778" Participant="me013">
And tha-
</Segment>
<Segment StartTime="1008.549" EndTime="1012.046" Participant="me013">
And that's what we were getting those numbers from. Right.
</Segment>
<Segment StartTime="1010.322" EndTime="1011.054" Participant="fe016">
Yes.
</Segment>
<Segment StartTime="1011.577" EndTime="1017.126" Participant="fe016">
Tho- good - the good numbers. The bad numbers were from <Pause/> the segments where there was overlap.
</Segment>
<Segment StartTime="1015.966" EndTime="1020.376" Participant="me013">
<VocalSound Description="inbreath"/> Well, we could <Emphasis> start </Emphasis> with the good ones. But anyway - so I think that we should try it once with <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1017.897" EndTime="1018.648" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="1020.376" EndTime="1026.290" Participant="me013">
the same conditions that were used to create <Emphasis> those, </Emphasis> and in those same segments just use one of the P_Z_Ms.
</Segment>
<Segment StartTime="1021.965" EndTime="1025.262" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1025.591" EndTime="1027.808" Participant="fe016">
Right. So we - we can do <Emphasis> that. </Emphasis> Yeah.
</Segment>
<Segment StartTime="1027.220" EndTime="1033.937" Participant="me013">
And then, you know, I mean, the thing is if we were getting, uh - what, thirty-five, forty percent, something like that on - on that particular set,
</Segment>
<Segment StartTime="1030.071" EndTime="1049.900" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1030.154" EndTime="1032.971" Participant="mn017">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="1032.585" EndTime="1051.894" Participant="mn005">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1033.937" EndTime="1034.599" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1034.599" EndTime="1039.426" Participant="me013">
uh, does it go to seventy or eighty? Or, does it use up so much memory we can't decode it?
</Segment>
<Segment StartTime="1037.303" EndTime="1037.936" Participant="fe016">
Right.
</Segment>
<Segment StartTime="1037.436" EndTime="1043.449" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1039.426" EndTime="1039.986" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1039.657" EndTime="1046.294" Participant="fe016">
It might also depend on <Emphasis> which </Emphasis> speaker th- it is and how close they are to the P_Z_M? I don't know how different they are from each other.
</Segment>
<Segment StartTime="1039.986" EndTime="1040.790" Participant="me013">
Uh -
</Segment>
<Segment StartTime="1044.575" EndTime="1048.980" Participant="mn017">
You want to probably choose the P_Z_M channel that is closest to the speaker.
</Segment>
<Segment StartTime="1046.789" EndTime="1047.975" Participant="fe016">
To be best -
</Segment>
<Segment StartTime="1047.864" EndTime="1051.420" Participant="me011">
<VocalSound Description="inbreath"/> For this particular digit ones, I just picked that one.
</Segment>
<Segment StartTime="1048.201" EndTime="1048.755" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1048.786" EndTime="1049.202" Participant="fe016">
f-
</Segment>
<Segment StartTime="1049.270" EndTime="1050.300" Participant="me013">
Well -
</Segment>
<Segment StartTime="1051.256" EndTime="1054.381" Participant="fe016">
O_K. So we would then use that one, too, or - ?
</Segment>
<Segment StartTime="1051.660" EndTime="1052.680" Participant="mn017">
Oh, O_K.
</Segment>
<Segment StartTime="1051.838" EndTime="1052.473" Participant="me011">
So -
</Segment>
<Segment StartTime="1052.580" EndTime="1054.283" Participant="me013">
<VocalSound Description="inbreath"/> This is kind of central.
</Segment>
<Segment StartTime="1054.283" EndTime="1063.144" Participant="me013">
You know, it's - so i- but I would - I'd pick that one. It'll be less good for some people than for other, but I - I'd like to see it on the same - exact same data set that - that we did the other thing on. Right?
</Segment>
<Segment StartTime="1054.456" EndTime="1055.010" Participant="mn005">
<VocalSound Description="cough"/>
</Segment>
<Segment StartTime="1058.703" EndTime="1059.750" Participant="me011">
Actually -
</Segment>
<Segment StartTime="1061.389" EndTime="1066.022" Participant="me011">
I sh- actually <Emphasis> should've </Emphasis> picked a different one, because <Pause/> that <Emphasis> could </Emphasis> be why the P_D_A is worse.
</Segment>
<Segment StartTime="1066.595" EndTime="1069.749" Participant="me011">
Because it's further away from most of the people reading digits.
</Segment>
<Segment StartTime="1067.191" EndTime="1068.437" Participant="mn014">
It's further away. Yeah.
</Segment>
<Segment StartTime="1069.859" EndTime="1070.175" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1070.489" EndTime="1071.698" Participant="me013">
That's probably <Emphasis> one </Emphasis> of the reasons.
</Segment>
<Segment StartTime="1070.524" EndTime="1071.075" Participant="fe008">
Hmm.
</Segment>
<Segment StartTime="1071.698" EndTime="1074.111" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1072.232" EndTime="1072.972" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1072.699" EndTime="1073.546" Participant="fe008">
Mm-hmm.
</Segment>
<Segment StartTime="1073.099" EndTime="1077.314" Participant="fe016">
Well, yeah. You could look at, I guess, <Emphasis> that </Emphasis> P_Z_M or something.
</Segment>
<Segment StartTime="1075.721" EndTime="1076.670" Participant="me011">
Yep.
</Segment>
<Segment StartTime="1076.959" EndTime="1080.899" Participant="me013">
<VocalSound Description="inbreath"/> But the other is, it's very, uh - I mean, even though there's - I'm sure the f- f-
</Segment>
<Segment StartTime="1080.899" EndTime="1085.543" Participant="me013">
the - the S_R_I, uh, front-end has some kind of pre-emphasis, it's - it's, uh -
</Segment>
<Segment StartTime="1084.604" EndTime="1085.358" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="1085.543" EndTime="1092.353" Participant="me013">
<VocalSound Description="inbreath"/> still, th- it's picking up <Emphasis> lots </Emphasis> of low-frequency energy. So, even discriminating against it, I'm sure some of it's getting through.
</Segment>
<Segment StartTime="1093.284" EndTime="1093.877" Participant="me013">
Um.
</Segment>
<Segment StartTime="1094.470" EndTime="1097.103" Participant="me013">
But, yeah, you're right. Prob- a part of it is just the distance.
</Segment>
<Segment StartTime="1097.871" EndTime="1099.770" Participant="fe016">
And aren't these pretty bad microphones?
</Segment>
<Segment StartTime="1099.460" EndTime="1101.578" Participant="me013">
<VocalSound Description="inbreath"/> Well, they're <Emphasis> bad. </Emphasis>
</Segment>
<Segment StartTime="1099.537" EndTime="1099.952" Participant="me011">
Yep.
</Segment>
<Segment StartTime="1100.205" EndTime="1100.897" Participant="fe016">
I mean -
</Segment>
<Segment StartTime="1101.578" EndTime="1105.050" Participant="me013">
<VocalSound Description="laugh"/> But, I mean, if you listen to it, it sounds O_K. You know?
</Segment>
<Segment StartTime="1101.759" EndTime="1103.796" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1101.916" EndTime="1102.786" Participant="me001">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="1106.565" EndTime="1107.910" Participant="me011">
Yeah. When you <Emphasis> listen </Emphasis> to it,
</Segment>
<Segment StartTime="1106.719" EndTime="1107.500" Participant="me013">
u- <Uncertain> Yeah. </Uncertain> <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="1108.642" EndTime="1115.230" Participant="me011">
uh, the P_Z_M and the P_D_A - Yeah, th- the P_D_A has higher sound floor but not by a lot. It's really pretty -
</Segment>
<Segment StartTime="1116.574" EndTime="1118.745" Participant="me011">
uh, pretty much the same.
</Segment>
<Segment StartTime="1116.750" EndTime="1119.776" Participant="fe016">
I just remember you saying you got them to be cheap on purpose.
</Segment>
<Segment StartTime="1120.076" EndTime="1120.412" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1120.172" EndTime="1122.110" Participant="fe016">
Cheap in terms of their quality. So.
</Segment>
<Segment StartTime="1122.153" EndTime="1125.400" Participant="me013">
Well, they're <Pause/> twenty-five cents or so. Yeah. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1122.196" EndTime="1126.694" Participant="me011">
Th- we wanted them to be - to be typical of what would be in a P_D_A. So they are -
</Segment>
<Segment StartTime="1126.205" EndTime="1126.798" Participant="fe016">
Mm-hmm. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="1127.040" EndTime="1130.413" Participant="me011">
they're not the P_Z_M three hundred dollar type. They're the twenty-five cent,
</Segment>
<Segment StartTime="1130.413" EndTime="1131.514" Participant="me011">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1130.624" EndTime="1130.947" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1131.514" EndTime="1133.591" Participant="me011">
buy <Pronounce Pronunciation="&quot;em&quot;"> them </Pronounce> in packs of thousand type. <VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="1133.220" EndTime="1133.735" Participant="me001">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="1133.232" EndTime="1133.786" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="1133.262" EndTime="1133.935" Participant="mn017">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="1133.590" EndTime="1134.124" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1134.098" EndTime="1134.591" Participant="fe016">
I see.
</Segment>
<Segment StartTime="1134.221" EndTime="1138.780" Participant="me013">
But, I mean, the thing is people use those little mikes for <Emphasis> everything </Emphasis> because they're <Emphasis> really </Emphasis> not <Emphasis> bad. </Emphasis>
</Segment>
<Segment StartTime="1134.570" EndTime="1135.171" Participant="mn005">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="1136.617" EndTime="1137.730" Participant="me011">
Everything.
</Segment>
<Segment StartTime="1138.780" EndTime="1140.113" Participant="me013">
I mean, if you're not
</Segment>
<Segment StartTime="1139.032" EndTime="1139.677" Participant="fe016">
Mm-hmm. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="1140.113" EndTime="1140.747" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1140.747" EndTime="1148.107" Participant="me013">
doing something <Emphasis> ridiculous </Emphasis> like feeding it to a speech recognizer, they - they - <VocalSound Description="breath-laugh"/> they - you know, you can hear the sou- hear the sounds just fine. <Comment Description="while laughing"/> You know, it's -
</Segment>
<Segment StartTime="1143.348" EndTime="1144.950" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1143.602" EndTime="1144.887" Participant="me026">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1143.605" EndTime="1145.379" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1143.730" EndTime="1145.753" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1143.737" EndTime="1145.090" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1146.453" EndTime="1148.560" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1147.324" EndTime="1148.033" Participant="fe016">
Right.
</Segment>
<Segment StartTime="1148.107" EndTime="1148.595" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1148.595" EndTime="1154.391" Participant="me013">
They - I mean, i- it's more or less the same <Emphasis> principles </Emphasis> as these other mikes are built under, it's just that <Pause/> there's less quality control.
</Segment>
<Segment StartTime="1154.391" EndTime="1156.191" Participant="me013">
They just, you know, churn <Pronounce Pronunciation="&quot;em&quot;"> them </Pronounce> out and don't check <Pronounce Pronunciation="&quot;em&quot;"> them. </Pronounce>
</Segment>
<Segment StartTime="1156.191" EndTime="1158.040" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1158.040" EndTime="1158.657" Participant="me013">
Um.
</Segment>
<Segment StartTime="1160.087" EndTime="1160.780" Participant="me013">
So.
</Segment>
<Segment StartTime="1161.610" EndTime="1162.697" Participant="mn005">
<VocalSound Description="cough"/>
</Segment>
<Segment StartTime="1162.760" EndTime="1167.200" Participant="me013">
So that was - Yeah. So that was i- interesting result. So like I said, the front-end guys are
</Segment>
<Segment StartTime="1168.120" EndTime="1171.310" Participant="me013">
very much interested in - in this is as - as well and
</Segment>
<Segment StartTime="1171.038" EndTime="1175.152" Participant="mn017">
So - so, but where is this now? I mean, what's - where do we go from here? I mean,
</Segment>
<Segment StartTime="1174.821" EndTime="1176.696" Participant="me011">
Yeah. That was gonna be my question.
</Segment>
<Segment StartTime="1175.152" EndTime="1180.334" Participant="mn017">
we - so we have a - we have a - a system that works pretty well but it's not, you know, the system that
</Segment>
<Segment StartTime="1180.334" EndTime="1184.745" Participant="mn017">
people here are <Emphasis> used </Emphasis> to using - to working with. So what - what do we do now?
</Segment>
<Segment StartTime="1181.887" EndTime="1182.540" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1183.588" EndTime="1189.820" Participant="me013">
Well, I think what we wanna do is we want to - eh, and we've talked about this in other <Pause/> contexts - we want to <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1190.560" EndTime="1193.409" Participant="me013">
have the ability to feed it different features.
</Segment>
<Segment StartTime="1193.843" EndTime="1195.014" Participant="mn017">
Mm-hmm. O_K.
</Segment>
<Segment StartTime="1194.369" EndTime="1196.090" Participant="me013">
And then, um, <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1195.799" EndTime="1199.696" Participant="mn005">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1196.090" EndTime="1201.061" Participant="me013">
from the point of view of the front-end research, it would be s- uh, substituting for H_T_K.
</Segment>
<Segment StartTime="1196.456" EndTime="1197.168" Participant="mn017">
<VocalSound Description="clears throat"/> <NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1201.185" EndTime="1201.776" Participant="mn017">
O_K.
</Segment>
<Segment StartTime="1202.162" EndTime="1208.123" Participant="me013">
I think that's the key thing. And then if we can feed it different features, then we can try all the different things that we're trying there. <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1203.291" EndTime="1204.041" Participant="mn017">
O_K.
</Segment>
<Segment StartTime="1207.709" EndTime="1208.203" Participant="mn017">
Alright.
</Segment>
<Segment StartTime="1208.123" EndTime="1209.491" Participant="me013">
And then, um,
</Segment>
<Segment StartTime="1209.928" EndTime="1212.205" Participant="me013">
uh, also Dave is - is thinking about
</Segment>
<Segment StartTime="1212.848" EndTime="1215.316" Participant="me013">
using the data in different ways, uh, to
</Segment>
<Segment StartTime="1214.782" EndTime="1215.432" Participant="mn017">
Mm-hmm. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="1215.316" EndTime="1215.881" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1215.881" EndTime="1216.544" Participant="me013">
um,
</Segment>
<Segment StartTime="1216.999" EndTime="1217.793" Participant="me013">
uh,
</Segment>
<Segment StartTime="1217.793" EndTime="1224.747" Participant="me013">
explicitly work on reverberation starting with some techniques that some other people have <Pause/> found somewhat useful, and - Yeah.
</Segment>
<Segment StartTime="1223.122" EndTime="1223.665" Participant="mn017">
O_K.
</Segment>
<Segment StartTime="1224.060" EndTime="1227.380" Participant="mn017">
So - so the key <Pause/> thing that's missing here is basically the
</Segment>
<Segment StartTime="1227.895" EndTime="1228.969" Participant="mn017">
ability to feed,
</Segment>
<Segment StartTime="1229.755" EndTime="1234.460" Participant="mn017">
you know, other features <VocalSound Description="outbreath"/> i- into the recognizer and also then to train the system.
</Segment>
<Segment StartTime="1232.825" EndTime="1233.249" Participant="me013">
Right.
</Segment>
<Segment StartTime="1234.418" EndTime="1234.853" Participant="me013">
Right.
</Segment>
<Segment StartTime="1235.461" EndTime="1241.260" Participant="mn017">
O_K. And, uh, es- I don't know when Chuck will be back but that's exactly what he - he's gonna -
</Segment>
<Segment StartTime="1238.591" EndTime="1239.362" Participant="me013">
<VocalSound Description="mouth"/> H- h-
</Segment>
<Segment StartTime="1239.362" EndTime="1245.680" Participant="me013">
He's - he's <Emphasis> sort </Emphasis> of back, but he drove for fourteen hours an- and wasn't gonna make it in today. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1242.840" EndTime="1244.641" Participant="mn017">
Oh, O_K. So,
</Segment>
<Segment StartTime="1245.386" EndTime="1249.084" Participant="mn017">
I think that's one of the things that he said he would be working on. Um.
</Segment>
<Segment StartTime="1250.610" EndTime="1253.562" Participant="mn017">
Just sort of t- to make sure that <Pause/> we can do that and -
</Segment>
<Segment StartTime="1250.760" EndTime="1251.143" Participant="me011">
Yeah.
</Segment>
<Segment StartTime="1253.507" EndTime="1253.995" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1254.283" EndTime="1255.062" Participant="mn017">
Um.
</Segment>
<Segment StartTime="1260.537" EndTime="1260.938" Participant="me013">
Right.
</Segment>
<Segment StartTime="1261.493" EndTime="1267.132" Participant="mn017">
It's - uh, I mean, the - the front-end is f- i- tha- that's in the S_R_I recognizer is very nice in that it
</Segment>
<Segment StartTime="1261.722" EndTime="1263.900" Participant="mn005">
<VocalSound Description="cough"/>
</Segment>
<Segment StartTime="1267.550" EndTime="1276.435" Participant="mn017">
does a lot of things on the fly but it unfortunately <Pause/> is not <Pause/> designed and, um - <VocalSound Description="mouth"/> like the, uh, ICSI system is, where you can feed it
</Segment>
<Segment StartTime="1276.880" EndTime="1279.401" Participant="mn017">
from a pipeline of - of the command. So,
</Segment>
<Segment StartTime="1279.880" EndTime="1283.550" Participant="mn017">
the - what that means probably for the foreseeable future is that
</Segment>
<Segment StartTime="1283.976" EndTime="1286.634" Participant="mn017">
you have to, uh, dump out, um -
</Segment>
<Segment StartTime="1285.315" EndTime="1286.010" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1287.110" EndTime="1290.050" Participant="mn017">
you know, if you want to use some new features, you have to dump them into
</Segment>
<Segment StartTime="1290.472" EndTime="1291.530" Participant="mn017">
individual files
</Segment>
<Segment StartTime="1291.981" EndTime="1294.424" Participant="mn017">
and <Pause/> give those files to the recognizer.
</Segment>
<Segment StartTime="1294.010" EndTime="1295.499" Participant="me011">
We do - we tend to do that anyway.
</Segment>
<Segment StartTime="1295.500" EndTime="1296.309" Participant="mn017">
O_K.
</Segment>
<Segment StartTime="1295.695" EndTime="1298.788" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1296.307" EndTime="1304.642" Participant="me011">
Oh. So, although you - you can pipe it as well, we tend to do it that way because that way you can concentrate on one block and not keep re-doing it over and over.
</Segment>
<Segment StartTime="1300.093" EndTime="1303.820" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1303.989" EndTime="1304.721" Participant="mn017">
Oh, O_K.
</Segment>
<Segment StartTime="1305.394" EndTime="1307.993" Participant="me001">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="1307.308" EndTime="1307.808" Participant="mn017">
Alright.
</Segment>
<Segment StartTime="1307.414" EndTime="1307.861" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1309.115" EndTime="1310.223" Participant="me013">
Yeah. So I've - I -
</Segment>
<Segment StartTime="1309.790" EndTime="1313.460" Participant="me011">
So tha- that's exactly what the P_file <Pause/> is for.
</Segment>
<Segment StartTime="1312.972" EndTime="1313.744" Participant="me013">
Yeah. <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1313.230" EndTime="1324.629" Participant="mn017">
<VocalSound Description="mouth"/> Yeah, the - the - the cumbersome thing is - is, um - is that you actually have to dump out little - little files. So for each segment that you want to recognize <VocalSound Description="breath"/> you have to <Pause/> dump out <Pause/> a separate file.
</Segment>
<Segment StartTime="1313.846" EndTime="1314.262" Participant="fe016">
Uh -
</Segment>
<Segment StartTime="1324.160" EndTime="1329.145" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1324.440" EndTime="1325.440" Participant="me011">
Uh-huh.
</Segment>
<Segment StartTime="1325.530" EndTime="1331.387" Participant="mn017">
Just like i- th- like th- as if there were these waveform segments, but instead you have sort of feature file segments.
</Segment>
<Segment StartTime="1333.155" EndTime="1335.125" Participant="mn017">
But, you know - <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="1336.050" EndTime="1338.028" Participant="mn017">
So. <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="1338.524" EndTime="1339.050" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="1339.050" EndTime="1340.086" Participant="me013">
Cool. <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="1339.360" EndTime="1340.313" Participant="mn017">
<VocalSound Description="clears throat"/>
</Segment>
<Segment StartTime="1340.086" EndTime="1344.161" Participant="me013">
<VocalSound Description="mouth"/> O_K. So the s- the - the next thing we had on the agenda was something about alignments?
</Segment>
<Segment StartTime="1344.922" EndTime="1345.532" Participant="fe016">
Oh.
</Segment>
<Segment StartTime="1346.224" EndTime="1347.525" Participant="fe016">
Yes, we have -
</Segment>
<Segment StartTime="1348.237" EndTime="1350.096" Participant="fe016">
I don't know, did <Emphasis> you </Emphasis> wanna talk about it, or - ?
</Segment>
<Segment StartTime="1350.096" EndTime="1352.944" Participant="fe016">
I can give a - I was just telling this to <Emphasis> Jane </Emphasis> and - and -
</Segment>
<Segment StartTime="1353.300" EndTime="1353.894" Participant="fe016">
W- we -
</Segment>
<Segment StartTime="1354.665" EndTime="1356.109" Participant="fe016">
we were able to get some
</Segment>
<Segment StartTime="1356.742" EndTime="1359.445" Participant="fe016">
definite improvement on the forced alignments by
</Segment>
<Segment StartTime="1360.058" EndTime="1365.775" Participant="fe016">
<Emphasis> looking </Emphasis> at them first and then realizing the kinds of errors <Pause/> that were occurring and
</Segment>
<Segment StartTime="1366.309" EndTime="1366.870" Participant="fe016">
um,
</Segment>
<Segment StartTime="1367.147" EndTime="1373.971" Participant="fe016">
some of the errors occurring very frequently are just things like the first word being moved to as early as possible in the recognition, which is a
</Segment>
<Segment StartTime="1368.828" EndTime="1370.460" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1374.723" EndTime="1375.296" Participant="fe016">
um,
</Segment>
<Segment StartTime="1375.672" EndTime="1379.291" Participant="fe016">
I think was <Emphasis> both </Emphasis> a - a pruning <Pause/> problem <Emphasis> and </Emphasis>
</Segment>
<Segment StartTime="1379.845" EndTime="1380.746" Participant="fe016">
possibly a
</Segment>
<Segment StartTime="1381.358" EndTime="1382.268" Participant="fe016">
problem with
</Segment>
<Segment StartTime="1382.625" EndTime="1385.418" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1382.711" EndTime="1387.647" Participant="fe016">
needing constraints on word locations. And so we tried both of these <Uncertain> st- </Uncertain> things. We tried saying -
</Segment>
<Segment StartTime="1388.145" EndTime="1389.091" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1388.329" EndTime="1390.930" Participant="fe016">
I don't know, I got this <VocalSound Description="outbreath"/> whacky idea that -
</Segment>
<Segment StartTime="1391.360" EndTime="1396.439" Participant="fe016">
just from looking at the data, that when people talk <Pause/> their words are usually chunked together.
</Segment>
<Segment StartTime="1397.237" EndTime="1399.025" Participant="fe016">
It's not that they say one word
</Segment>
<Segment StartTime="1399.440" EndTime="1401.030" Participant="fe016">
and then there's a bunch of words together.
</Segment>
<Segment StartTime="1401.030" EndTime="1405.002" Participant="fe016">
They're <Comment Description="likely a speech error"/> <Emphasis> might </Emphasis> say one word and then another word far away if they were doing just backchannels?
</Segment>
<Segment StartTime="1405.002" EndTime="1412.520" Participant="fe016">
But in general, if there's, like, five or six words and one word's far <Emphasis> away </Emphasis> from it, that's probably <Emphasis> wrong </Emphasis> on average. So,
</Segment>
<Segment StartTime="1413.052" EndTime="1413.645" Participant="fe016">
um -
</Segment>
<Segment StartTime="1414.484" EndTime="1418.917" Participant="fe016">
And then also, ca- the pruning, of course, was too - too severe.
</Segment>
<Segment StartTime="1417.869" EndTime="1422.595" Participant="mn017">
<VocalSound Description="inbreath"/> So that's actually interesting. The pruning was the same value that we used for recognition.
</Segment>
<Segment StartTime="1423.208" EndTime="1428.768" Participant="mn017">
And we had lowered that - we had used <Emphasis> tighter </Emphasis> pruning after Liz ran some experiments showing that,
</Segment>
<Segment StartTime="1427.092" EndTime="1427.823" Participant="fe016">
<VocalSound Description="clears throat"/>
</Segment>
<Segment StartTime="1428.768" EndTime="1432.219" Participant="mn017">
you know, it runs <Emphasis> slower </Emphasis> and there's no real difference in -
</Segment>
<Segment StartTime="1431.138" EndTime="1432.008" Participant="me011">
No gain.
</Segment>
<Segment StartTime="1431.293" EndTime="1435.800" Participant="fe016">
Actually it was <Emphasis> better </Emphasis> with - slightly better or about th- it was the <Emphasis> same </Emphasis> with tighter pruning.
</Segment>
<Segment StartTime="1433.007" EndTime="1433.953" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1433.610" EndTime="1434.088" Participant="mn017">
Right.
</Segment>
<Segment StartTime="1434.964" EndTime="1439.318" Participant="mn017">
So for free recognition, this - the lower pruning value is <Emphasis> better. </Emphasis> You -
</Segment>
<Segment StartTime="1438.987" EndTime="1440.904" Participant="fe016">
It's probably cuz the recognition's just bad en-
</Segment>
<Segment StartTime="1441.379" EndTime="1445.693" Participant="fe016">
at a point where it's <Emphasis> bad </Emphasis> enough that - that you don't lose anything.
</Segment>
<Segment StartTime="1442.087" EndTime="1442.502" Participant="mn017">
Correct.
</Segment>
<Segment StartTime="1442.128" EndTime="1442.834" Participant="me013">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="1443.544" EndTime="1444.090" Participant="mn017">
Right.
</Segment>
<Segment StartTime="1444.958" EndTime="1450.947" Participant="mn017">
Um, but it turned out for - for - to get accurate <Emphasis> alignments </Emphasis> it was really important to open up the pruning
</Segment>
<Segment StartTime="1445.690" EndTime="1446.350" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1446.245" EndTime="1447.223" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1447.597" EndTime="1448.199" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1451.168" EndTime="1451.702" Participant="fe016">
Right. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="1451.280" EndTime="1452.488" Participant="mn017">
significantly.
</Segment>
<Segment StartTime="1452.289" EndTime="1452.744" Participant="me013">
Hmm.
</Segment>
<Segment StartTime="1452.488" EndTime="1460.362" Participant="mn017">
Um <Pause/> because otherwise it would sort of do greedy alignment, um, in regions where there was no real speech <Emphasis> yet </Emphasis> from the foreground speaker.
</Segment>
<Segment StartTime="1460.379" EndTime="1461.057" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="1461.179" EndTime="1466.878" Participant="mn017">
Um, <VocalSound Description="mouth"/> so that was one big factor that helped improve things and then the other thing was that,
</Segment>
<Segment StartTime="1467.728" EndTime="1470.502" Participant="mn017">
you know, as Liz said the - we f- enforce
</Segment>
<Segment StartTime="1471.140" EndTime="1472.178" Participant="mn017">
the fact that,
</Segment>
<Segment StartTime="1472.650" EndTime="1475.661" Participant="mn017">
uh, the foreground speech has to be continuous. It cannot be -
</Segment>
<Segment StartTime="1476.025" EndTime="1478.039" Participant="mn017">
you cannot have a background speech
</Segment>
<Segment StartTime="1478.554" EndTime="1484.116" Participant="mn017">
hypothesis in the <Emphasis> middle </Emphasis> of the foreground speech. You can only have background speech at the beginning and the end.
</Segment>
<Segment StartTime="1484.370" EndTime="1489.812" Participant="fe016">
Yeah. I mean, yeah, it isn't <Emphasis> always </Emphasis> true, and I think what we really want is some <Emphasis> clever </Emphasis> way to do this, where,
</Segment>
<Segment StartTime="1488.171" EndTime="1488.824" Participant="mn017">
<VocalSound Description="clears throat"/>
</Segment>
<Segment StartTime="1490.346" EndTime="1495.214" Participant="fe016">
um, you know, from the <Emphasis> data </Emphasis> or from maybe some hand-corrected alignments from transcribers that
</Segment>
<Segment StartTime="1495.649" EndTime="1496.390" Participant="fe016">
things like
</Segment>
<Segment StartTime="1496.834" EndTime="1504.348" Participant="fe016">
words that <Emphasis> do </Emphasis> occur just by themselves <Pause/> a- alone, like backchannels or something that we <Emphasis> did </Emphasis> allow to have background speech around it -
</Segment>
<Segment StartTime="1501.363" EndTime="1501.831" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1504.348" EndTime="1507.910" Participant="fe016">
<Emphasis> those </Emphasis> would be able to do that, but the rest would be constrained.
</Segment>
<Segment StartTime="1504.404" EndTime="1513.949" Participant="fe008">
<NonVocalSound Description="cell phone ringing"/> Sorry. <Comment Description="quietly"/> <NonVocalSound Description="zipper"/> <NonVocalSound Description="cell phone ringing"/>
</Segment>
<Segment StartTime="1508.634" EndTime="1510.716" Participant="fe016">
So, I think we have a version that's
</Segment>
<Segment StartTime="1511.131" EndTime="1514.929" Participant="fe016">
pretty good for the native speakers. I don't know yet about the non-native speakers.
</Segment>
<Segment StartTime="1515.641" EndTime="1517.050" Participant="fe016">
And, um,
</Segment>
<Segment StartTime="1517.757" EndTime="1523.822" Participant="fe016">
we basically also made noise models for the different - sort of grouped some of the <Pause/> mouth noises together.
</Segment>
<Segment StartTime="1524.385" EndTime="1526.917" Participant="fe016">
Um, so, and then there's a background speech model.
</Segment>
<Segment StartTime="1527.434" EndTime="1528.364" Participant="fe016">
And we also -
</Segment>
<Segment StartTime="1528.364" EndTime="1532.254" Participant="fe016">
There was some neat - or, interesting cases, like there's one meeting where,
</Segment>
<Segment StartTime="1532.254" EndTime="1535.714" Participant="fe016">
<VocalSound Description="inbreath"/> um, Jose's giving a presentation and he's talking about,
</Segment>
<Segment StartTime="1536.151" EndTime="1540.481" Participant="fe016">
um, the word "mixed <Pause/> signal" and someone didn't understand,
</Segment>
<Segment StartTime="1540.481" EndTime="1542.304" Participant="fe016">
uh, that you were saying "<Emphasis> mixed </Emphasis>" -
</Segment>
<Segment StartTime="1542.304" EndTime="1543.036" Participant="fe016">
I think, Morgan.
</Segment>
<Segment StartTime="1542.360" EndTime="1544.355" Participant="mn005">
<VocalSound Description="laugh"/> Yeah, yeah.
</Segment>
<Segment StartTime="1542.418" EndTime="1544.259" Participant="me013">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="1543.036" EndTime="1543.438" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1543.203" EndTime="1544.152" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1543.438" EndTime="1545.339" Participant="fe016">
And so your speech -ch
</Segment>
<Segment StartTime="1545.788" EndTime="1546.401" Participant="mn017">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="1545.940" EndTime="1549.137" Participant="fe016">
was s- saying something about mixed signal. And the next turn
</Segment>
<Segment StartTime="1549.552" EndTime="1550.393" Participant="fe016">
was
</Segment>
<Segment StartTime="1550.960" EndTime="1557.879" Participant="fe016">
a <Emphasis> lot </Emphasis> of people saying "mixed", like "he means mixed signal" or "I think it's mixed". And the word "mixed" in this <Emphasis> segment </Emphasis> occurs, like, a bunch of times.
</Segment>
<Segment StartTime="1554.959" EndTime="1555.695" Participant="mn005">
Sh-
</Segment>
<Segment StartTime="1557.879" EndTime="1558.331" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1558.331" EndTime="1565.928" Participant="fe016">
And Chuck's on the <Emphasis> lapel </Emphasis> here, and he also says "mixed" but it's at the <Emphasis> last </Emphasis> one, and of course the aligner th- aligns it <Emphasis> everywhere </Emphasis> else to <Emphasis> everybody </Emphasis> else's "mixed",
</Segment>
<Segment StartTime="1564.847" EndTime="1566.730" Participant="mn005">
<VocalSound Description="laugh"/> Yeah.
</Segment>
<Segment StartTime="1565.294" EndTime="1566.323" Participant="me011">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="1565.846" EndTime="1566.597" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1565.928" EndTime="1566.354" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1566.354" EndTime="1568.667" Participant="fe016">
cuz there's no adaptation yet. So there's - <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1568.875" EndTime="1569.508" Participant="mn017">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="1569.023" EndTime="1570.361" Participant="fe016">
I think there's some issues about -
</Segment>
<Segment StartTime="1570.764" EndTime="1571.188" Participant="fe016">
u-
</Segment>
<Segment StartTime="1571.188" EndTime="1572.623" Participant="fe016">
We probably want to adapt
</Segment>
<Segment StartTime="1571.345" EndTime="1571.781" Participant="mn014">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="1573.098" EndTime="1580.157" Participant="fe016">
at least the foreground speaker. But, I guess Andreas tried adapting both the foreground and a background generic speaker, and that's actually a little bit
</Segment>
<Segment StartTime="1574.154" EndTime="1574.930" Participant="mn017">
<VocalSound Description="clears throat"/>
</Segment>
<Segment StartTime="1581.210" EndTime="1591.656" Participant="fe016">
of a f- funky model. Like, it gives you some weird alignments, just because often the <Emphasis> background </Emphasis> speakers match better to the foreground than the foreground speaker. So there's some things <Emphasis> there, </Emphasis> especially when you get lots of the same words,
</Segment>
<Segment StartTime="1584.529" EndTime="1585.281" Participant="mn017">
<VocalSound Description="mouth"/> Oh -
</Segment>
<Segment StartTime="1587.626" EndTime="1588.062" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1587.858" EndTime="1588.860" Participant="mn005">
Oh.
</Segment>
<Segment StartTime="1592.308" EndTime="1593.456" Participant="fe016">
uh, occurring in the -
</Segment>
<Segment StartTime="1592.369" EndTime="1592.962" Participant="mn017">
Well, the -
</Segment>
<Segment StartTime="1593.652" EndTime="1595.480" Participant="mn017">
I - I think you can do better by
</Segment>
<Segment StartTime="1595.480" EndTime="1595.714" Participant="mn017">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="1596.228" EndTime="1598.958" Participant="mn017">
uh, cloning - so we have a reject phone.
</Segment>
<Segment StartTime="1599.393" EndTime="1605.694" Participant="mn017">
And you - and what we wanted to try with - you know, once we have this paper written and have a little more time,
</Segment>
<Segment StartTime="1605.694" EndTime="1606.228" Participant="mn017">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1606.228" EndTime="1611.364" Participant="mn017">
uh, t- cloning that reject model and then one copy of it would be adapted to the foreground speaker
</Segment>
<Segment StartTime="1611.364" EndTime="1617.710" Participant="mn017">
to capture the rejects in the <Emphasis> foreground, </Emphasis> like fragments and stuff, and the other copy would be adapted to the background speaker.
</Segment>
<Segment StartTime="1618.109" EndTime="1618.405" Participant="mn017">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="1618.396" EndTime="1620.396" Participant="fe016">
Right. I mean, in general we actually -
</Segment>
<Segment StartTime="1618.970" EndTime="1619.474" Participant="mn017">
And -
</Segment>
<Segment StartTime="1620.396" EndTime="1624.336" Participant="fe016">
Right now the words like <Pause/> partial words are <Pause/> reject models
</Segment>
<Segment StartTime="1624.332" EndTime="1624.905" Participant="mn017">
Mm-hmm. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="1624.810" EndTime="1629.693" Participant="fe016">
and you normally allow those to match to any word. But then the background speech was also a reject model,
</Segment>
<Segment StartTime="1629.693" EndTime="1630.040" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1630.040" EndTime="1636.360" Participant="fe016">
and so this constraint of not allowing rejects in between - you know, it needs to differentiate between the two. So just sort of working through a bunch of
</Segment>
<Segment StartTime="1631.660" EndTime="1632.669" Participant="mn005">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="1634.941" EndTime="1635.388" Participant="mn017">
Right.
</Segment>
<Segment StartTime="1637.039" EndTime="1638.600" Participant="fe016">
debugging kinds of issues.
</Segment>
<Segment StartTime="1638.815" EndTime="1639.767" Participant="fe008">
<NonVocalSound Description="mechanical sound"/>
</Segment>
<Segment StartTime="1639.167" EndTime="1641.445" Participant="fe016">
And <Emphasis> another </Emphasis> one is turns, like people starting with
</Segment>
<Segment StartTime="1641.445" EndTime="1646.020" Participant="fe016">
<VocalSound Description="inbreath"/> "well I think" and someone else is <Pause/> "well how about". So the word "well" is in this -
</Segment>
<Segment StartTime="1646.375" EndTime="1647.593" Participant="fe016">
in this <Pause/> segment
</Segment>
<Segment StartTime="1648.074" EndTime="1653.038" Participant="fe016">
multiple times, and as soon as it occurs usually the aligner will try to align it to the first person who says it.
</Segment>
<Segment StartTime="1653.038" EndTime="1653.424" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1653.424" EndTime="1654.129" Participant="fe016">
But then that
</Segment>
<Segment StartTime="1654.584" EndTime="1655.997" Participant="fe016">
constraint of sort of -
</Segment>
<Segment StartTime="1656.366" EndTime="1660.617" Participant="fe016">
uh, <Emphasis> proximity </Emphasis> constraint will push it over to the person who <Emphasis> really </Emphasis> said it in general.
</Segment>
<Segment StartTime="1658.987" EndTime="1659.323" Participant="me011">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="1659.739" EndTime="1665.430" Participant="me011">
Is the proximity constraint a <Emphasis> hard </Emphasis> constraint, or did you do some sort of probabilistic weighting distance, or - ?
</Segment>
<Segment StartTime="1662.390" EndTime="1662.957" Participant="fe016">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="1663.477" EndTime="1667.183" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1663.632" EndTime="1665.841" Participant="mn017">
We - we didn't - No. We - w-
</Segment>
<Segment StartTime="1663.860" EndTime="1667.479" Participant="fe016">
<VocalSound Description="laugh"/> Right now it's a kluge. <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1666.266" EndTime="1666.997" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1666.269" EndTime="1670.898" Participant="mn017">
O_K. We - it's straightforward to actually just have a - a penalty that doesn't
</Segment>
<Segment StartTime="1670.898" EndTime="1673.350" Participant="mn017">
completely disallows it but discourages it.
</Segment>
<Segment StartTime="1673.775" EndTime="1679.050" Participant="mn017">
But, um, we just didn't have time to play with, you know, tuning yet another - yet another parameter. <VocalSound Description="breath"/>
</Segment>
<Segment StartTime="1676.829" EndTime="1678.411" Participant="me011">
The ve- level. Yeah.
</Segment>
<Segment StartTime="1678.496" EndTime="1679.167" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="1679.050" EndTime="1684.013" Participant="mn017">
And really the reason we can't do it is just that we don't have a - we don't have ground truth for these. So,
</Segment>
<Segment StartTime="1684.013" EndTime="1684.315" Participant="mn017">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1684.315" EndTime="1692.817" Participant="mn017">
we would need a hand-marked, um, <VocalSound Description="mouth"/> word-level alignments or at least sort of the boundaries of the speech betw- you know, between the speakers.
</Segment>
<Segment StartTime="1692.817" EndTime="1693.290" Participant="mn017">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1693.290" EndTime="1693.920" Participant="mn017">
Um,
</Segment>
<Segment StartTime="1694.240" EndTime="1696.304" Participant="mn017">
and then use that as a reference and tune
</Segment>
<Segment StartTime="1696.620" EndTime="1701.490" Participant="mn017">
the parameters of the - of the model, uh, to op- to get the best <Pause/> performance.
</Segment>
<Segment StartTime="1699.243" EndTime="1699.777" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="1699.672" EndTime="1700.470" Participant="fe008">
<VocalSound Description="mouth"/> <VocalSound Description="inbreath"/> <Comment Description="as if to speak"/>
</Segment>
<Segment StartTime="1700.412" EndTime="1701.008" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="1700.772" EndTime="1707.220" Participant="me013">
G- given - I - I mean, I wa- I wa- I was gonna ask you anyway, uh, how you assessed that things were better.
</Segment>
<Segment StartTime="1703.140" EndTime="1703.879" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="1707.572" EndTime="1711.767" Participant="fe016">
I looked at them. I spent <Emphasis> two </Emphasis> <Emphasis> days </Emphasis> - um, in <Emphasis> Waves </Emphasis> -
</Segment>
<Segment StartTime="1708.901" EndTime="1709.356" Participant="me013">
O_K.
</Segment>
<Segment StartTime="1709.574" EndTime="1712.800" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1709.794" EndTime="1713.554" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1710.431" EndTime="1711.687" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1710.445" EndTime="1711.518" Participant="mn017">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1711.507" EndTime="1712.390" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1711.767" EndTime="1713.676" Participant="fe016">
Oh, it was <Emphasis> painful </Emphasis> because
</Segment>
<Segment StartTime="1713.075" EndTime="1713.517" Participant="me026">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1713.676" EndTime="1714.062" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1714.062" EndTime="1716.989" Participant="fe016">
the thing is, you know the alignments share a lot in common, so -
</Segment>
<Segment StartTime="1716.989" EndTime="1721.671" Participant="fe016">
And you're - yo- you're looking at these segments where there's a <Emphasis> lot </Emphasis> of speech. I mean, a lot of them have a lot of words.
</Segment>
<Segment StartTime="1721.709" EndTime="1722.160" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1722.045" EndTime="1725.150" Participant="fe016">
Not by <Emphasis> every </Emphasis> speaker but by <Emphasis> some </Emphasis> speaker there's a lot of words.
</Segment>
<Segment StartTime="1724.448" EndTime="1725.002" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1724.793" EndTime="1725.107" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1724.955" EndTime="1727.810" Participant="mn005">
Ju- <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1725.150" EndTime="1728.060" Participant="fe016">
No, not - I mean that if you look at the individual
</Segment>
<Segment StartTime="1725.712" EndTime="1728.160" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1726.159" EndTime="1728.540" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1727.071" EndTime="1727.921" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1727.536" EndTime="1728.367" Participant="mn017">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="1728.877" EndTime="1729.688" Participant="me013">
<VocalSound Description="inbreath"/> Yeah.
</Segment>
<Segment StartTime="1728.947" EndTime="1736.152" Participant="fe016">
segments from just one <Emphasis> person </Emphasis> you don't see a lot of words, but <Emphasis> altogether </Emphasis> you'll see a lot of words up there. And so the <Emphasis> reject </Emphasis> is also mapping and <Emphasis> pauses </Emphasis> -
</Segment>
<Segment StartTime="1729.159" EndTime="1729.555" Participant="mn014">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="1730.839" EndTime="1731.611" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="1731.123" EndTime="1731.473" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1733.473" EndTime="1733.852" Participant="mn014">
Yeah. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="1736.152" EndTime="1736.646" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1736.646" EndTime="1740.099" Participant="fe016">
So I looked at <Pronounce Pronunciation="&quot;em&quot;"> them </Pronounce> all in Waves and just lined up all the alignments,
</Segment>
<Segment StartTime="1740.099" EndTime="1746.447" Participant="fe016">
and, at first it sort of looked like a <Emphasis> mess </Emphasis> and then the more I looked at it, I thought "O_K, well it's moving these words leftward and -"
</Segment>
<Segment StartTime="1740.434" EndTime="1743.599" Participant="mn005">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1746.724" EndTime="1750.603" Participant="fe016">
You know, it wasn't <Emphasis> that </Emphasis> bad. It was just doing certain things wrong. So - But, I don't,
</Segment>
<Segment StartTime="1751.610" EndTime="1757.873" Participant="fe016">
you know, have <Emphasis> time </Emphasis> to l- <Comment Description="while laughing"/> to look at all of them and it would be <Emphasis> really </Emphasis> useful to have, like, a - a transcriber who could use Waves,
</Segment>
<Segment StartTime="1752.576" EndTime="1754.467" Participant="mn017">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="1757.873" EndTime="1758.350" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1758.350" EndTime="1761.047" Participant="fe016">
um, just mark, like, the beginning and end
</Segment>
<Segment StartTime="1761.462" EndTime="1766.889" Participant="fe016">
of the foreground speaker's <Emphasis> real </Emphasis> <Emphasis> words </Emphasis> - like, the beginning of the first word, the end of the last word - and then we could,
</Segment>
<Segment StartTime="1763.449" EndTime="1764.235" Participant="fe008">
<VocalSound Description="mouth"/> <VocalSound Description="inbreath"/> <Comment Description="as if to speak"/>
</Segment>
<Segment StartTime="1764.389" EndTime="1765.507" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1765.507" EndTime="1766.515" Participant="fe008">
Yeah.
</Segment>
<Segment StartTime="1766.515" EndTime="1769.996" Participant="fe008">
I - O_K. I have to ask you something, which is, first of all, um,
</Segment>
<Segment StartTime="1767.547" EndTime="1769.195" Participant="fe016">
you know, do some adjustments. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="1769.996" EndTime="1776.524" Participant="fe008">
is i- does it <Emphasis> have </Emphasis> to be <Emphasis> Waves? </Emphasis> Because if we could benefit from what <Emphasis> you </Emphasis> did, incorporate that into the present transcripts, <Comment Description="embedded"/> that would help.
</Segment>
<Segment StartTime="1772.340" EndTime="1773.132" Participant="mn017">
No.
</Segment>
<Segment StartTime="1776.524" EndTime="1780.361" Participant="fe008">
<VocalSound Description="inbreath"/> And then, um, the <Emphasis> other </Emphasis> thing is, I believe that I <Emphasis> did </Emphasis> hand-
</Segment>
<Segment StartTime="1780.361" EndTime="1791.226" Participant="fe008">
So. One of these transcripts was gone over by a transcriber and then I hand-marked it myself so that we <Emphasis> do </Emphasis> have, uh, the beginning and ending of individual <Emphasis> utterances. </Emphasis>
</Segment>
<Segment StartTime="1791.328" EndTime="1791.820" Participant="mn017">
Mm-hmm. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="1791.652" EndTime="1799.877" Participant="fe008">
Um, I didn't do it <Emphasis> word </Emphasis> level, but - but in terms - So I - so for - for one of the N_S_<Emphasis> A </Emphasis> groups. And <Emphasis> also </Emphasis> I went back to the original one that I <Emphasis> first </Emphasis> transcribed and - and did it
</Segment>
<Segment StartTime="1794.207" EndTime="1794.678" Participant="fe016">
Mm-hmm.
</Segment>
<Segment StartTime="1799.877" EndTime="1804.347" Participant="fe008">
w- uh, w- uh, utterance by utterance for <Emphasis> that </Emphasis> particular one. So I think you <Emphasis> do </Emphasis> have -
</Segment>
<Segment StartTime="1804.347" EndTime="1809.189" Participant="fe008">
if that's a sufficient unit, I think that you <Emphasis> do </Emphasis> have hand-marking for <Emphasis> that. </Emphasis> But it'd be wonderful to be able to
</Segment>
<Segment StartTime="1809.189" EndTime="1811.369" Participant="fe008">
<VocalSound Description="inbreath"/> benefit from your Waves stuff.
</Segment>
<Segment StartTime="1809.396" EndTime="1809.832" Participant="fe016">
Mm-hmm. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="1809.850" EndTime="1812.263" Participant="mn017">
We don't care what - what tool you use.
</Segment>
<Segment StartTime="1812.299" EndTime="1814.514" Participant="fe008">
O_K. I used it in Transcriber and it's - it's in the -
</Segment>
<Segment StartTime="1812.343" EndTime="1816.615" Participant="fe016">
Yeah. I mean, if - if you can, um - if you wanna - well, Jane and I were -
</Segment>
<Segment StartTime="1812.639" EndTime="1813.410" Participant="mn017">
U- uh -
</Segment>
<Segment StartTime="1816.615" EndTime="1825.355" Participant="fe016">
just in terms of the <Emphasis> tool, </Emphasis> talking about this. I guess <Emphasis> Sue </Emphasis> had had some <Pause/> reactions. You know, interface-wise if you're looking at speech, you wanna be able to know really where the words are. And so,
</Segment>
<Segment StartTime="1821.260" EndTime="1826.482" Participant="mn005">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1823.199" EndTime="1824.000" Participant="mn017">
<VocalSound Description="clears throat"/>
</Segment>
<Segment StartTime="1824.599" EndTime="1826.713" Participant="fe008">
Yeah, that's right. Middle of the word, or -
</Segment>
<Segment StartTime="1825.355" EndTime="1825.730" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1825.418" EndTime="1825.952" Participant="mn017">
<VocalSound Description="clears throat"/>
</Segment>
<Segment StartTime="1825.730" EndTime="1831.150" Participant="fe016">
we can give you some examples of sort of what this output looks like, um, and see if you can in-
</Segment>
<Segment StartTime="1831.970" EndTime="1834.965" Participant="fe016">
maybe incorporate it into the Transcriber tool some way, <Comment Description="rising intonation"/> or -
</Segment>
<Segment StartTime="1833.482" EndTime="1834.333" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1834.728" EndTime="1839.418" Participant="fe008">
Well, I th- I'm thinking just ch- e- e- incorporating it into the <Emphasis> representation. </Emphasis> I mean, if it's - if it's -
</Segment>
<Segment StartTime="1835.971" EndTime="1836.790" Participant="fe016">
Um.
</Segment>
<Segment StartTime="1838.544" EndTime="1840.651" Participant="fe016">
You mean like - <Emphasis> Yeah, </Emphasis> word start insights.
</Segment>
<Segment StartTime="1839.418" EndTime="1845.631" Participant="fe008">
if you have start points, if you have, like, <Emphasis> time </Emphasis> tags, which is what I <Emphasis> assume. </Emphasis> Isn't that what - what you - ? Well, see, Adam would be -
</Segment>
<Segment StartTime="1841.691" EndTime="1842.410" Participant="fe016">
Right.
</Segment>
<Segment StartTime="1844.490" EndTime="1844.846" Participant="mn017">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="1844.846" EndTime="1847.571" Participant="mn017">
Yeah, whatever you use. I mean, we convert it to this format
</Segment>
<Segment StartTime="1846.053" EndTime="1846.621" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="1848.157" EndTime="1851.340" Participant="mn017">
that the, um, NIST scoring tool
</Segment>
<Segment StartTime="1851.900" EndTime="1853.062" Participant="mn017">
unders- uh, C_T_M.
</Segment>
<Segment StartTime="1853.062" EndTime="1857.520" Participant="mn017">
Conversation Time-Marked file. And - and then that's the - that's what the -
</Segment>
<Segment StartTime="1857.170" EndTime="1860.000" Participant="me011">
I think Transcriber, uh, outputs C_T_M.
</Segment>
<Segment StartTime="1860.495" EndTime="1862.089" Participant="fe008">
If it - ? O_K. So you would know this more than I would.
</Segment>
<Segment StartTime="1860.941" EndTime="1861.940" Participant="me011">
I <Emphasis> think </Emphasis> so.
</Segment>
<Segment StartTime="1861.260" EndTime="1862.808" Participant="fe016">
Yeah. <Pause/> So, I mean -
</Segment>
<Segment StartTime="1861.456" EndTime="1862.294" Participant="mn017">
<VocalSound Description="breath"/> Right. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="1862.089" EndTime="1871.091" Participant="fe008">
It <Emphasis> seems </Emphasis> like she - if she's g- if she's moving <Emphasis> time </Emphasis> marks around, since our representation in Transcriber uses <Emphasis> time </Emphasis> marks, it seems like there should be <Emphasis> some </Emphasis> way of - of <Emphasis> using </Emphasis> that - <Emphasis> benefitting </Emphasis> from that.
</Segment>
<Segment StartTime="1867.683" EndTime="1868.177" Participant="me011">
Right.
</Segment>
<Segment StartTime="1870.219" EndTime="1875.439" Participant="fe016">
Yeah, it wou- the advantage would just be that when you brought up a bin you <Emphasis> would </Emphasis> be able - if you were zoomed in <Emphasis> enough </Emphasis>
</Segment>
<Segment StartTime="1875.439" EndTime="1880.618" Participant="fe016">
in Transcriber to see all the words, you would be able to, like, have the words sort of located in time,
</Segment>
<Segment StartTime="1875.945" EndTime="1876.562" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="1879.861" EndTime="1880.770" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1880.618" EndTime="1882.400" Participant="fe016">
if you wanted to do that. So.
</Segment>
<Segment StartTime="1880.637" EndTime="1884.634" Participant="me013">
<VocalSound Description="mouth"/> So - so if we e- e- even just had a - a - It sounds like w- we -
</Segment>
<Segment StartTime="1884.634" EndTime="1885.406" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1885.406" EndTime="1886.810" Participant="me013">
we almost <Emphasis> do. </Emphasis>
</Segment>
<Segment StartTime="1886.810" EndTime="1888.397" Participant="me013">
Uh, if we - We have two. Yeah.
</Segment>
<Segment StartTime="1887.011" EndTime="1887.871" Participant="fe008">
We have two.
</Segment>
<Segment StartTime="1888.397" EndTime="1889.871" Participant="me013">
<VocalSound Description="inbreath"/> Just ha- uh,
</Segment>
<Segment StartTime="1890.211" EndTime="1891.844" Participant="me013">
trying out <Pause/> the alignment
</Segment>
<Segment StartTime="1891.844" EndTime="1892.319" Participant="me013">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="1892.316" EndTime="1892.717" Participant="fe016">
Mm-hmm.
</Segment>
<Segment StartTime="1892.319" EndTime="1897.123" Participant="me013">
procedure that you have on <Emphasis> that </Emphasis> you could actually get something, um - uh, uh, get an objective measure.
</Segment>
<Segment StartTime="1897.123" EndTime="1897.936" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1897.495" EndTime="1898.182" Participant="mn017">
Mm-hmm. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="1897.936" EndTime="1898.707" Participant="me013">
Uh -
</Segment>
<Segment StartTime="1898.896" EndTime="1901.086" Participant="fe016">
You mean on - on the hand-marked, um -
</Segment>
<Segment StartTime="1901.086" EndTime="1908.525" Participant="fe016">
<VocalSound Description="inbreath"/> So we - we only r- hav- I only <Emphasis> looked </Emphasis> at actually alignments from <Emphasis> one </Emphasis> meeting that we chose, I think M_R four, just randomly, um -
</Segment>
<Segment StartTime="1901.556" EndTime="1902.290" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1908.525" EndTime="1909.097" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1909.620" EndTime="1910.220" Participant="fe016">
And -
</Segment>
<Segment StartTime="1910.610" EndTime="1915.338" Participant="mn017">
Actually, <Emphasis> not </Emphasis> randomly. We knew - we knew that it had these insertion errors from - Yeah.
</Segment>
<Segment StartTime="1911.596" EndTime="1912.793" Participant="fe016">
Not randomly -
</Segment>
<Segment StartTime="1913.294" EndTime="1917.076" Participant="fe016">
It had sort of <Pause/> average recognition performance in a bunch of speakers and
</Segment>
<Segment StartTime="1915.736" EndTime="1916.550" Participant="mn017">
Yeah.
</Segment>
<Segment StartTime="1917.735" EndTime="1918.914" Participant="fe016">
it was a Meeting Recorder meeting.
</Segment>
<Segment StartTime="1918.914" EndTime="1924.582" Participant="fe016">
Um. But, yeah, we should try to <Emphasis> use </Emphasis> what you have. I <Emphasis> did </Emphasis> re-run recognition on your new version of
</Segment>
<Segment StartTime="1924.213" EndTime="1924.589" Participant="me013">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1924.638" EndTime="1926.715" Participant="fe008">
Oh, good. Good!
</Segment>
<Segment StartTime="1925.144" EndTime="1929.634" Participant="fe016">
M_R <Emphasis> one. </Emphasis> I - I mean the - the one with Dan <Pause/> Ellis in it <VocalSound Description="laugh"/> and Eric.
</Segment>
<Segment StartTime="1925.637" EndTime="1926.132" Participant="me013">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1927.506" EndTime="1929.147" Participant="fe008">
Uh-huh. Yeah, exactly. Yeah. Yeah.
</Segment>
<Segment StartTime="1928.445" EndTime="1929.236" Participant="mn014">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="1929.576" EndTime="1932.160" Participant="me001">
<VocalSound Description="mouth"/> I don't think that was the <Emphasis> new </Emphasis> version.
</Segment>
<Segment StartTime="1929.634" EndTime="1930.405" Participant="fe016">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="1930.405" EndTime="1930.781" Participant="fe016">
Um -
</Segment>
<Segment StartTime="1931.144" EndTime="1938.954" Participant="fe016">
That - Yeah, actually it wasn't the new new, it was the medium new. But - but we would - we should do the - the <Emphasis> latest </Emphasis> version. It was the one from last week.
</Segment>
<Segment StartTime="1932.545" EndTime="1933.052" Participant="fe008">
O_K.
</Segment>
<Segment StartTime="1933.287" EndTime="1934.058" Participant="mn005">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="1933.796" EndTime="1934.251" Participant="fe008">
O_K.
</Segment>
<Segment StartTime="1933.919" EndTime="1937.251" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1934.915" EndTime="1935.676" Participant="mn017">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="1935.058" EndTime="1935.880" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1935.516" EndTime="1936.861" Participant="mn014">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="1936.030" EndTime="1942.650" Participant="me001">
Yeah. You - did you adjust the - the utterance times, um, for each channel?
</Segment>
<Segment StartTime="1940.837" EndTime="1941.549" Participant="fe008">
Yes.
</Segment>
<Segment StartTime="1942.569" EndTime="1945.507" Participant="fe008">
Yes, I did. And <Emphasis> furthermore, </Emphasis> I found that there were a certain number where -
</Segment>
<Segment StartTime="1945.507" EndTime="1954.013" Participant="fe008">
<VocalSound Description="breath"/> not - not a <Emphasis> lot, </Emphasis> but several times I actually <VocalSound Description="breath"/> <Emphasis> moved </Emphasis> an utterance from <VocalSound Description="inbreath"/> Adam's channel to <Emphasis> Dan's </Emphasis> or from <Emphasis> Dan's </Emphasis> to <Emphasis> Adam's. </Emphasis> So there was some speaker identif- <Comment Description="abandons as obvious"/>
</Segment>
<Segment StartTime="1954.013" EndTime="1957.594" Participant="fe008">
<VocalSound Description="inbreath"/> And the <Emphasis> reason </Emphasis> was because <VocalSound Description="breath"/> I transcribed that at a point before -
</Segment>
<Segment StartTime="1956.591" EndTime="1957.125" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1957.594" EndTime="1958.029" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1958.029" EndTime="1961.095" Participant="fe008">
uh, before we had the multiple <Emphasis> audio </Emphasis> available f-
</Segment>
<Segment StartTime="1961.095" EndTime="1967.724" Participant="fe008">
so I couldn't <Emphasis> switch </Emphasis> between the <Pronounce> audio. </Pronounce> I - I transcribed it off of the mixed channel <Emphasis> entirely, </Emphasis> which meant in overlaps, I was at a - at a terrific disadvantage.
</Segment>
<Segment StartTime="1966.637" EndTime="1968.261" Participant="fe016">
Right. Right.
</Segment>
<Segment StartTime="1967.724" EndTime="1974.112" Participant="fe008">
<VocalSound Description="inbreath"/> In <Emphasis> addition </Emphasis> it was before the <Emphasis> channelized, </Emphasis> uh, possibility was there. And <Emphasis> finally </Emphasis> I did it using the speakers of my, um - <VocalSound Description="breath"/>
</Segment>
<Segment StartTime="1970.300" EndTime="1970.800" Participant="mn017">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="1974.112" EndTime="1977.541" Participant="fe008">
of - you know, off the C_P_U on my - on my machine cuz I didn't have a headphone.
</Segment>
<Segment StartTime="1977.541" EndTime="1978.444" Participant="fe008">
So it <Uncertain NumSyllables="2"> @@ </Uncertain>, like, I mean -
</Segment>
<Segment StartTime="1977.749" EndTime="1978.422" Participant="fe016">
Right.
</Segment>
<Segment StartTime="1978.439" EndTime="1979.389" Participant="mn005">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="1978.444" EndTime="1981.511" Participant="fe008">
<VocalSound Description="inbreath"/> Yeah, I - I mean, i- in retrospect <VocalSound Description="laugh"/> it would've been good to ha-
</Segment>
<Segment StartTime="1978.613" EndTime="1979.048" Participant="mn017">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="1981.511" EndTime="1983.490" Participant="fe008">
<VocalSound Description="inbreath"/> have got- I should've gotten a headphone.
</Segment>
<Segment StartTime="1983.490" EndTime="1987.770" Participant="fe008">
But in <Emphasis> any </Emphasis> case, um, thi- this is - this was transcribed in a - in a,
</Segment>
<Segment StartTime="1987.770" EndTime="1996.586" Participant="fe008">
<VocalSound Description="inbreath"/> uh, less optimal way than - than the ones that came after it, and I was able to - you know, an- and <Emphasis> this </Emphasis> meant that there were some speaker identif- identifications which were
</Segment>
<Segment StartTime="1995.773" EndTime="2000.204" Participant="me001">
Well, I know there were some speaker labeling problems, um, after interruptions. Is that what you're referring to?
</Segment>
<Segment StartTime="1997.061" EndTime="1998.149" Participant="fe008">
changes. Yeah.
</Segment>
<Segment StartTime="1998.572" EndTime="1999.343" Participant="fe008">
Fixed that.
</Segment>
<Segment StartTime="1999.778" EndTime="2001.249" Participant="fe008">
<VocalSound Description="inbreath"/> Oh, well -
</Segment>
<Segment StartTime="2000.115" EndTime="2002.509" Participant="mn005">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="2000.204" EndTime="2004.532" Participant="me001">
I mean, cuz there's this one instance when, for example, you're running down the stairs. I remember this meeting really well.
</Segment>
<Segment StartTime="2003.906" EndTime="2005.020" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2004.260" EndTime="2005.132" Participant="mn014">
Yeah. <VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2004.436" EndTime="2010.147" Participant="fe016">
<VocalSound Description="laugh"/> Don - Don has had - <VocalSound Description="laugh"/> He knows - he can just read it like a play. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2004.519" EndTime="2008.052" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2004.532" EndTime="2005.843" Participant="me001">
<VocalSound Description="inbreath"/> Right. <Uncertain> It's a </Uncertain> -
</Segment>
<Segment StartTime="2004.926" EndTime="2006.322" Participant="me026">
<VocalSound Description="laugh"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2005.843" EndTime="2008.739" Participant="me001">
Yeah, I've - I've - I'm very well acquainted with this meeting.
</Segment>
<Segment StartTime="2007.212" EndTime="2008.260" Participant="me026">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="2008.423" EndTime="2010.325" Participant="mn017">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2008.950" EndTime="2009.840" Participant="mn014">
<VocalSound Description="breath"/> Yeah. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="2009.146" EndTime="2010.669" Participant="me001">
Yeah, I can s- <VocalSound Description="breath"/>
</Segment>
<Segment StartTime="2009.556" EndTime="2014.462" Participant="me026">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2010.147" EndTime="2011.870" Participant="fe016">
"And then <Emphasis> she </Emphasis> said, and then <Emphasis> he </Emphasis> said." <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2010.986" EndTime="2017.633" Participant="me001">
Yeah, I know it by heart. So, um, <VocalSound Description="mouth"/> there's one point when you're running down the stairs. Right? And, like, there's an interruption.
</Segment>
<Segment StartTime="2011.150" EndTime="2013.630" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2012.292" EndTime="2013.870" Participant="mn017">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2012.662" EndTime="2013.038" Participant="me013">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2012.670" EndTime="2013.656" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2015.033" EndTime="2016.675" Participant="fe008">
Uh-oh. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2017.633" EndTime="2023.312" Participant="me001">
You interrupt somebody, but then there's no line after that. For example, there's no speaker identification after that line.
</Segment>
<Segment StartTime="2022.214" EndTime="2022.807" Participant="fe008">
Uh-huh. <Comment Description="while smiling"/>
</Segment>
<Segment StartTime="2023.678" EndTime="2027.349" Participant="me001">
Is that what you're talking about? Or were there mislabelings as far as, like, the a- Adam was - ?
</Segment>
<Segment StartTime="2025.715" EndTime="2026.961" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2026.961" EndTime="2029.324" Participant="fe008">
That was fixed, um, before -
</Segment>
<Segment StartTime="2029.324" EndTime="2032.726" Participant="fe008">
i- i- i- I think I- I think I understood that pretty - Thank you for mentioning. Yeah, no, tha- that -
</Segment>
<Segment StartTime="2029.643" EndTime="2032.313" Participant="me001">
Yeah. Cuz I thought I let you know about that. Yeah.
</Segment>
<Segment StartTime="2032.726" EndTime="2036.129" Participant="fe008">
That I think went away a couple of versions ago, but it's good to know.
</Segment>
<Segment StartTime="2034.509" EndTime="2038.210" Participant="me001">
O_K. But you're actually saying that certain, uh, speakers were mis- mis-identified.
</Segment>
<Segment StartTime="2036.129" EndTime="2037.039" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2037.039" EndTime="2043.034" Participant="fe008">
<Emphasis> Yeah. </Emphasis> So, with - under - um, uh, listening to the mixed channel, there were times when, as surprising as that is,
</Segment>
<Segment StartTime="2043.034" EndTime="2048.673" Participant="fe008">
I got Adam's voice confused with Dan's and vice versa - not for <Emphasis> long </Emphasis> utterances, but jus- just a couple of places,
</Segment>
<Segment StartTime="2044.199" EndTime="2044.931" Participant="me001">
O_K.
</Segment>
<Segment StartTime="2046.070" EndTime="2046.543" Participant="me001">
O_K.
</Segment>
<Segment StartTime="2046.912" EndTime="2047.564" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="2048.673" EndTime="2053.080" Participant="fe008">
<VocalSound Description="inbreath"/> and embedde- embedded in overlaps. The <Emphasis> other </Emphasis> thing that was w- interesting to me was
</Segment>
<Segment StartTime="2048.882" EndTime="2049.439" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="2053.080" EndTime="2058.420" Participant="fe008">
that I picked up a lot of, um, <Emphasis> backchannels </Emphasis> which were hidden in the mixed signal, which, you know, I mean, you c- not - not too surprising.
</Segment>
<Segment StartTime="2056.866" EndTime="2057.576" Participant="fe016">
Right.
</Segment>
<Segment StartTime="2058.420" EndTime="2059.803" Participant="fe008">
<VocalSound Description="inbreath"/> But the <Emphasis> other </Emphasis> thing that -
</Segment>
<Segment StartTime="2058.826" EndTime="2059.162" Participant="me001">
<NonVocalSound Description="noise"/>
</Segment>
<Segment StartTime="2059.803" EndTime="2062.928" Participant="fe008">
I - I <Emphasis> hadn't </Emphasis> thought about this, but I thou- I wanted to raise this when you were -
</Segment>
<Segment StartTime="2062.928" EndTime="2066.983" Participant="fe008">
uh, with respect to also a <Emphasis> strategy </Emphasis> which might help with the alignments <Emphasis> potentially, </Emphasis> but that's -
</Segment>
<Segment StartTime="2066.983" EndTime="2067.754" Participant="fe008">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2067.754" EndTime="2069.435" Participant="fe008">
When I was looking at these backchannels,
</Segment>
<Segment StartTime="2069.910" EndTime="2071.230" Participant="fe008">
they were turning up usually -
</Segment>
<Segment StartTime="2071.230" EndTime="2071.551" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2071.551" EndTime="2075.982" Participant="fe008">
<Emphasis> very </Emphasis> often in - w- well, I won't say "usually" - but anyway, very often, I picked them up in a channel
</Segment>
<Segment StartTime="2075.982" EndTime="2076.545" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2076.545" EndTime="2082.985" Participant="fe008">
w- which was the person who had asked a question. <VocalSound Description="laugh"/> S- so, like, someone says "an- and have you done the so-and-so?"
</Segment>
<Segment StartTime="2082.985" EndTime="2083.424" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2083.424" EndTime="2088.321" Participant="fe008">
And then there would be <Emphasis> backchannels, </Emphasis> but it would be the person who asked the <Emphasis> question. </Emphasis> Other people weren't really <Emphasis> doing </Emphasis> much backchanneling.
</Segment>
<Segment StartTime="2088.321" EndTime="2090.771" Participant="fe008">
<VocalSound Description="inbreath"/> And, you know, sometimes you have the - Yeah, uh-huh.
</Segment>
<Segment StartTime="2089.510" EndTime="2090.855" Participant="fe016">
Well, that's interesting.
</Segment>
<Segment StartTime="2090.771" EndTime="2095.201" Participant="fe008">
<VocalSound Description="inbreath"/> I mean, i- it wouldn't be <Emphasis> perfect, </Emphasis> but - but it <Emphasis> does </Emphasis> seem more natural to give a backchannel when -
</Segment>
<Segment StartTime="2091.363" EndTime="2092.002" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="2093.354" EndTime="2094.986" Participant="fe016">
No, that's really interesting.
</Segment>
<Segment StartTime="2095.537" EndTime="2100.622" Participant="fe008">
when you're somehow involved in the <Emphasis> topic, </Emphasis> and the most natural way is for you to have <Emphasis> initiated </Emphasis> the topic by asking a question. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2097.390" EndTime="2097.998" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="2099.415" EndTime="2100.028" Participant="mn017">
Well,
</Segment>
<Segment StartTime="2100.622" EndTime="2103.134" Participant="fe008">
<VocalSound Description="breath-laugh"/> <Comment Description="some mike noise embedded"/>
</Segment>
<Segment StartTime="2100.888" EndTime="2102.389" Participant="fe016">
<VocalSound Description="inbreath"/> That's interesting.
</Segment>
<Segment StartTime="2100.950" EndTime="2111.580" Participant="mn017">
I think - No. I think it's - actually I think what's going on is <Emphasis> backchanneling </Emphasis> is something that happens in two-party conversations. And if you ask someone a question, you essentially initiating a little two-party conversation.
</Segment>
<Segment StartTime="2107.042" EndTime="2107.537" Participant="fe008">
Mm-hmm.
</Segment>
<Segment StartTime="2110.821" EndTime="2111.237" Participant="fe008">
Yeah. <Comment Description="whispered"/>
</Segment>
<Segment StartTime="2111.488" EndTime="2113.506" Participant="fe016">
Well, actu- Yeah, when we <Emphasis> looked </Emphasis> at this -
</Segment>
<Segment StartTime="2111.930" EndTime="2117.393" Participant="mn017">
So then you're - so and then you're expected to backchannel because the person is addressing <Emphasis> you </Emphasis> directly and not everybody.
</Segment>
<Segment StartTime="2111.988" EndTime="2112.779" Participant="fe008">
<Emphasis> Exactly. </Emphasis> <Comment Description="quietly"/>
</Segment>
<Segment StartTime="2115.390" EndTime="2117.052" Participant="fe008">
<VocalSound Description="inbreath"/> <Emphasis> Exactly. </Emphasis> <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2117.052" EndTime="2121.586" Participant="fe008">
Exactly my point. An- and so this is the expectation thing that - uh, uh, just the dyadic - But in <Emphasis> addition, </Emphasis>
</Segment>
<Segment StartTime="2117.781" EndTime="2119.230" Participant="mn017">
Yeah. Yeah. Right.
</Segment>
<Segment StartTime="2118.272" EndTime="2119.201" Participant="me013">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2118.590" EndTime="2119.065" Participant="fe016">
Mm-hmm.
</Segment>
<Segment StartTime="2119.745" EndTime="2120.482" Participant="mn017">
Right.
</Segment>
<Segment StartTime="2121.555" EndTime="2122.168" Participant="me013">
<VocalSound Description="inbreath"/> H-
</Segment>
<Segment StartTime="2121.586" EndTime="2125.582" Participant="fe008">
you know, if someone has <Emphasis> done </Emphasis> this analysis <Emphasis> himself </Emphasis> and <Emphasis> isn't </Emphasis> involved in the <Emphasis> dyad, </Emphasis>
</Segment>
<Segment StartTime="2125.582" EndTime="2130.863" Participant="fe008">
but they might also give backchannels to verify what - what the answer is that this - that the - the <Emphasis> answerer's </Emphasis> given -
</Segment>
<Segment StartTime="2129.811" EndTime="2131.849" Participant="me013">
I tell you, I say - I say "uh-huh" a <Emphasis> lot, </Emphasis>
</Segment>
<Segment StartTime="2129.845" EndTime="2131.065" Participant="fe016">
Right. It's -
</Segment>
<Segment StartTime="2131.353" EndTime="2135.929" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2131.474" EndTime="2131.989" Participant="fe008">
There you go.
</Segment>
<Segment StartTime="2131.740" EndTime="2134.340" Participant="fe016">
<VocalSound Description="inbreath"/> Well, but it's interesting cuz, uh -
</Segment>
<Segment StartTime="2131.849" EndTime="2139.207" Participant="me013">
<VocalSound Description="laugh"/> while people are talking to each other. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2131.989" EndTime="2133.492" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2132.267" EndTime="2133.830" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2134.620" EndTime="2137.206" Participant="fe008">
<Emphasis> There </Emphasis> you go. Yeah. Yeah.
</Segment>
<Segment StartTime="2135.180" EndTime="2135.793" Participant="me001">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2135.386" EndTime="2144.954" Participant="fe016">
But there are <Emphasis> fewer </Emphasis> - I think there are fewer "uh-huhs". I mean, just from - We were looking at word frequency lists to try to find the cases that we would <Emphasis> allow </Emphasis> to have the <Emphasis> reject </Emphasis> words in <Emphasis> between </Emphasis> in doing the <Emphasis> alignment. </Emphasis>
</Segment>
<Segment StartTime="2135.910" EndTime="2136.784" Participant="mn017">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2145.170" EndTime="2147.575" Participant="fe016">
You know the ones we wouldn't constrain to be next to the other words.
</Segment>
<Segment StartTime="2145.250" EndTime="2146.225" Participant="fe008">
<VocalSound Description="mouth"/> Oh, yeah.
</Segment>
<Segment StartTime="2147.575" EndTime="2148.260" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2148.260" EndTime="2148.913" Participant="fe016">
And
</Segment>
<Segment StartTime="2149.770" EndTime="2151.424" Participant="fe016">
"uh-huh" is not as frequent
</Segment>
<Segment StartTime="2151.990" EndTime="2156.555" Participant="fe016">
as it sort of would be in Switchboard, if you looked at just a word frequency list of one-word
</Segment>
<Segment StartTime="2154.369" EndTime="2157.051" Participant="mn017">
<VocalSound Description="outbreath"/> <VocalSound Description="clears throat"/>
</Segment>
<Segment StartTime="2156.930" EndTime="2160.889" Participant="fe016">
short utterances. And "<Emphasis> yeah </Emphasis>" is way up there, but not "uh-huh". And so I was thinking
</Segment>
<Segment StartTime="2160.889" EndTime="2161.530" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2161.530" EndTime="2171.507" Participant="fe016">
thi- it's not like <Pause/> you're being encouraged by everybody else to keep <Pause/> talking in the meeting. And uh, that's all, I- I'll stop there, cuz I- I think what you say makes a lot of sense. <Comment Description="while laughing"/> But it was sort of - <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2162.273" EndTime="2163.020" Participant="mn014">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2166.461" EndTime="2168.407" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2168.062" EndTime="2169.327" Participant="fe008">
<VocalSound Description="mouth"/> Well, that's right. And that would -
</Segment>
<Segment StartTime="2169.327" EndTime="2170.614" Participant="fe008">
<VocalSound Description="inbreath"/> Well, an-
</Segment>
<Segment StartTime="2170.614" EndTime="2175.084" Participant="fe008">
And what you say is the - is the re- uh, o- <Emphasis> other </Emphasis> side of this, which is that, you know, so
</Segment>
<Segment StartTime="2175.084" EndTime="2180.227" Participant="fe008">
th- there are lots of channels where you <Emphasis> don't </Emphasis> have these backchannels, w- when a question has been asked and - and these -
</Segment>
<Segment StartTime="2179.399" EndTime="2181.872" Participant="fe016">
Right. There's just probably less backchanneling
</Segment>
<Segment StartTime="2180.227" EndTime="2180.801" Participant="fe008">
Mm-hmm.
</Segment>
<Segment StartTime="2182.256" EndTime="2183.542" Participant="fe008">
So that's good news, really.
</Segment>
<Segment StartTime="2182.465" EndTime="2187.738" Participant="fe016">
in <Emphasis> general, </Emphasis> even if you consider <Emphasis> every </Emphasis> other person altogether one person
</Segment>
<Segment StartTime="2188.421" EndTime="2194.728" Participant="fe016">
in the meeting, but we'll find out anyway. We were - I guess the other thing we're - we're - I should say is that we're gonna, um
</Segment>
<Segment StartTime="2195.526" EndTime="2198.660" Participant="fe016">
try - compare this type of <Emphasis> overlap </Emphasis> analysis to <Emphasis> Switchboard, </Emphasis>
</Segment>
<Segment StartTime="2198.896" EndTime="2199.786" Participant="mn017">
And CallHome.
</Segment>
<Segment StartTime="2199.117" EndTime="2203.706" Participant="fe016">
where - and <Emphasis> CallHome, </Emphasis> where we <Emphasis> have </Emphasis> both sides, so that we can try to answer this question of, you know,
</Segment>
<Segment StartTime="2201.308" EndTime="2201.856" Participant="me011">
Mm-hmm.
</Segment>
<Segment StartTime="2201.332" EndTime="2202.206" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="2203.706" EndTime="2204.193" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2204.193" EndTime="2209.714" Participant="fe016">
is there really <Emphasis> more </Emphasis> <Emphasis> overlap </Emphasis> in <Emphasis> meetings </Emphasis> or is it just because we don't have the other channel in <Emphasis> Switchboard </Emphasis> and we don't know what people are doing.
</Segment>
<Segment StartTime="2209.714" EndTime="2211.494" Participant="fe016">
Try to create a paper out of that.
</Segment>
<Segment StartTime="2210.095" EndTime="2210.471" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2211.620" EndTime="2218.782" Participant="me013">
<VocalSound Description="inbreath"/> Yeah. I mean, y- y- you folks have probably <Pause/> already told me, but were - were you intending to do a Eurospeech submission, or - ?
</Segment>
<Segment StartTime="2212.502" EndTime="2213.215" Participant="fe016">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2212.937" EndTime="2213.679" CloseMic="false">
<NonVocalSound Description="someone knocks on wood, nonverbal gesture for good luck"/>
</Segment>
<Segment StartTime="2214.043" EndTime="2215.120" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2218.526" EndTime="2221.236" Participant="fe016">
<VocalSound Description="laugh"/> Um, you mean the one due <Emphasis> tomorrow? </Emphasis>
</Segment>
<Segment StartTime="2219.058" EndTime="2220.640" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2219.842" EndTime="2220.582" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2221.294" EndTime="2222.402" Participant="me013">
Yeah. <VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2221.746" EndTime="2222.259" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2221.928" EndTime="2224.302" Participant="fe016">
Yeah. Well, we're still, like, writing the scripts
</Segment>
<Segment StartTime="2224.920" EndTime="2228.080" Participant="fe016">
for doing the research, and we will - Yes, we're gonna try.
</Segment>
<Segment StartTime="2226.388" EndTime="2226.962" Participant="fe008">
Mm-hmm. <Comment Description="very quietly"/>
</Segment>
<Segment StartTime="2227.829" EndTime="2230.236" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2228.785" EndTime="2229.482" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2229.187" EndTime="2234.837" Participant="fe016">
<VocalSound Description="inbreath"/> And I was telling Don, do not <Pause/> take this as an example of how people should work. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2233.240" EndTime="2238.724" Participant="me013">
<VocalSound Description="laugh"/> Do as I say, don't do as I do. Yeah. <VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2234.042" EndTime="2235.269" Participant="me026">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2234.642" EndTime="2235.313" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2235.082" EndTime="2237.090" Participant="me001">
That's r- <Comment Description="whispered"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2235.380" EndTime="2237.000" Participant="fe016">
<VocalSound Description="breath-laugh"/> So, <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2236.536" EndTime="2237.226" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2237.000" EndTime="2240.181" Participant="fe016">
we will <Emphasis> try. </Emphasis> It'll probably be a little late, but I'm gonna try it.
</Segment>
<Segment StartTime="2237.165" EndTime="2237.919" Participant="me011">
<VocalSound Description="inbreath"/> Well -
</Segment>
<Segment StartTime="2239.000" EndTime="2243.926" Participant="me011">
It <Emphasis> is </Emphasis> different. In previous years, Eurospeech only had the <Emphasis> abstract </Emphasis> due by now, not the full paper.
</Segment>
<Segment StartTime="2243.003" EndTime="2243.852" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2243.248" EndTime="2243.790" Participant="me001">
Right.
</Segment>
<Segment StartTime="2243.729" EndTime="2244.203" Participant="fe016">
Right.
</Segment>
<Segment StartTime="2244.340" EndTime="2250.512" Participant="me011">
And so all our timing was off. I've given up on trying to do digits. I just don't think that what I have so far makes a Eurospeech paper.
</Segment>
<Segment StartTime="2251.088" EndTime="2253.927" Participant="fe016">
Well, I'm no- We may be in the same position, and I figured
</Segment>
<Segment StartTime="2253.927" EndTime="2254.589" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2254.589" EndTime="2261.672" Participant="fe016">
we'll <Emphasis> try, </Emphasis> because that'll at least get us to the point where we have - We have this really nice database format that Andreas and I were working out that -
</Segment>
<Segment StartTime="2260.191" EndTime="2263.850" Participant="me013">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2261.672" EndTime="2261.983" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2261.983" EndTime="2263.898" Participant="fe016">
It - it's not very <Emphasis> fancy. </Emphasis> It's just a
</Segment>
<Segment StartTime="2263.141" EndTime="2263.814" Participant="mn017">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2264.323" EndTime="2267.490" Participant="fe016">
ASCII line by line format, but it <Emphasis> does </Emphasis> give you information -
</Segment>
<Segment StartTime="2266.650" EndTime="2269.710" Participant="mn017">
It's the - it's the spurt format.
</Segment>
<Segment StartTime="2268.976" EndTime="2270.384" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2269.000" EndTime="2275.514" Participant="fe016">
It - Yeah, we're calling these "spurts" after Chafe. I was trying to find what's a word for <Pause/> a continuous region with <Pause/> pauses around it?
</Segment>
<Segment StartTime="2272.124" EndTime="2272.658" Participant="fe008">
Hmm.
</Segment>
<Segment StartTime="2274.297" EndTime="2274.693" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2275.543" EndTime="2280.065" Participant="me013">
Yeah. I know that th- the Telecom people use - use "spurt" for that. Yes.
</Segment>
<Segment StartTime="2275.571" EndTime="2275.993" Participant="fe008">
Good.
</Segment>
<Segment StartTime="2278.747" EndTime="2281.589" Participant="fe016">
They do? Oh! Oh. I would jus-
</Segment>
<Segment StartTime="2279.239" EndTime="2279.773" Participant="mn017">
Oh.
</Segment>
<Segment StartTime="2280.065" EndTime="2287.067" Participant="me013">
And that's - I mean, I - I was using that for a while when I was doing the rate of speech stuff, because I - because I looked up in some books and I found - O_K, I wanna find a <Emphasis> spurt </Emphasis>
</Segment>
<Segment StartTime="2283.828" EndTime="2285.012" Participant="fe016">
Ah, right!
</Segment>
<Segment StartTime="2287.067" EndTime="2287.603" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2287.384" EndTime="2289.661" Participant="fe016">
It's just, like, defined by the <Emphasis> acoustics. </Emphasis>
</Segment>
<Segment StartTime="2287.603" EndTime="2294.704" Participant="me013">
in which - and - an- because - cuz it's <Emphasis> another </Emphasis> question about how <Pause/> many pauses they put in <Emphasis> between </Emphasis> them. But how fast do they
</Segment>
<Segment StartTime="2291.190" EndTime="2292.218" Participant="me011">
Horrible.
</Segment>
<Segment StartTime="2292.515" EndTime="2293.111" Participant="fe016">
Right.
</Segment>
<Segment StartTime="2295.050" EndTime="2297.544" Participant="me013">
do <Pause/> the words within the spurt? Yeah.
</Segment>
<Segment StartTime="2296.491" EndTime="2297.091" Participant="me011">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2296.624" EndTime="2299.789" Participant="me001">
<VocalSound Description="mouth"/> <Uncertain> you know </Uncertain> "<Emphasis> Burst </Emphasis>" also? Isn't "burst" is used also?
</Segment>
<Segment StartTime="2296.649" EndTime="2298.924" Participant="fe016">
Right. Well, that's what we were calling
</Segment>
<Segment StartTime="2297.091" EndTime="2299.320" Participant="me011">
It's gonna - Burst.
</Segment>
<Segment StartTime="2299.450" EndTime="2300.605" Participant="fe016">
spurt, so -
</Segment>
<Segment StartTime="2300.520" EndTime="2307.000" Participant="me011">
Spurt has the horrible name overloading with other - with hardware at ICSI. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2303.329" EndTime="2307.425" Participant="mn005">
<VocalSound Description="laugh"/> Here <Uncertain NumSyllables="2"> @@ </Uncertain> - <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2303.879" EndTime="2306.998" Participant="me013">
<Emphasis> Here. </Emphasis> Just very locally, yeah. But - but that just -
</Segment>
<Segment StartTime="2306.000" EndTime="2311.030" Participant="fe016">
Well, well, <Emphasis> Chafe </Emphasis> had this wor- I think it was Chafe, or somebody had a - the word "spurt" originally, and so I -
</Segment>
<Segment StartTime="2308.481" EndTime="2309.015" Participant="fe008">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2311.488" EndTime="2312.073" Participant="fe008">
Actually -
</Segment>
<Segment StartTime="2311.554" EndTime="2314.074" Participant="fe016">
But tha- that's good to know. Was thi- it's Chafe?
</Segment>
<Segment StartTime="2312.073" EndTime="2312.951" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2313.465" EndTime="2319.241" Participant="fe008">
Well, see, I know S- Sue wrote about spurts of development. <Emphasis> But, </Emphasis> in any case, I think it's a good term, and, uh -
</Segment>
<Segment StartTime="2313.480" EndTime="2314.778" Participant="mn017">
So maybe we should talk -
</Segment>
<Segment StartTime="2315.201" EndTime="2316.784" Participant="fe016">
Maybe it was Sue - ? Y-
</Segment>
<Segment StartTime="2317.900" EndTime="2320.096" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2318.242" EndTime="2318.735" Participant="me011">
Hmm! <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2318.544" EndTime="2322.737" Participant="fe016">
So we have spurts and we have spurt-ify dot shell and spurt-ify <Pause/> <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2318.663" EndTime="2319.199" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="2319.241" EndTime="2321.358" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2321.358" EndTime="2324.433" Participant="fe008">
And ma- maybe - maybe Chafe did. I know - I know Ch- Chafe dealt with -
</Segment>
<Segment StartTime="2321.560" EndTime="2325.646" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2321.590" EndTime="2322.519" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="2321.622" EndTime="2323.303" Participant="mn017">
Uh. So s-
</Segment>
<Segment StartTime="2321.662" EndTime="2323.998" Participant="me001">
<VocalSound Description="laugh"/> That's cool.
</Segment>
<Segment StartTime="2321.664" EndTime="2324.663" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2322.737" EndTime="2325.730" Participant="fe016">
And then it's got all - it's a verb now. <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2324.272" EndTime="2325.795" Participant="mn017">
W- uh, w- <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2324.927" EndTime="2326.688" Participant="fe008">
Chafe speaks about <Emphasis> intonation </Emphasis> units.
</Segment>
<Segment StartTime="2324.940" EndTime="2325.785" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2325.432" EndTime="2326.127" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2326.684" EndTime="2327.564" Participant="fe016">
Yes. Right.
</Segment>
<Segment StartTime="2326.688" EndTime="2329.714" Participant="fe008">
But maybe he <Emphasis> speaks </Emphasis> about spurts as well and I just don't know. Yeah, go ahead.
</Segment>
<Segment StartTime="2327.452" EndTime="2327.926" Participant="mn017">
We-
</Segment>
<Segment StartTime="2329.160" EndTime="2335.404" Participant="mn017">
So what we're doing - uh, this - this is just - maybe someone has s- some - some ideas about how to do it better, but we -
</Segment>
<Segment StartTime="2329.438" EndTime="2331.060" Participant="me011">
I've heard "burst" also.
</Segment>
<Segment StartTime="2330.809" EndTime="2331.403" Participant="me001">
Mmm.
</Segment>
<Segment StartTime="2335.960" EndTime="2338.798" Participant="mn017">
So we're taking these, uh, alignments from the individual channels.
</Segment>
<Segment StartTime="2339.082" EndTime="2339.497" Participant="mn017">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2339.497" EndTime="2340.177" Participant="mn017">
We're -
</Segment>
<Segment StartTime="2340.754" EndTime="2349.067" Participant="mn017">
from each alignment we're producing, uh, one of these C_T_M files, which essentially has - it's just a linear sequence of words with the begin times for every word and the duration.
</Segment>
<Segment StartTime="2344.291" EndTime="2344.730" Participant="fe008">
Great.
</Segment>
<Segment StartTime="2350.100" EndTime="2351.660" Participant="mn017">
And - and - and of course -
</Segment>
<Segment StartTime="2350.555" EndTime="2353.568" Participant="fe016">
It looks like a Waves label file almost. Right? It's just -
</Segment>
<Segment StartTime="2352.672" EndTime="2358.806" Participant="mn017">
Right. But it has - one - the first column has the meeting <Emphasis> name, </Emphasis> so it could actually contain several meetings. Um.
</Segment>
<Segment StartTime="2359.332" EndTime="2360.704" Participant="mn017">
And the second column is the channel.
</Segment>
<Segment StartTime="2361.456" EndTime="2366.566" Participant="mn017">
Third column is the, um, <Emphasis> start </Emphasis> times of the words and the fourth column is the <Emphasis> duration </Emphasis> of the words.
</Segment>
<Segment StartTime="2366.566" EndTime="2367.098" Participant="mn017">
<VocalSound Description="breath"/> <VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2367.492" EndTime="2368.678" Participant="mn017">
And then we're,
</Segment>
<Segment StartTime="2369.410" EndTime="2369.700" Participant="mn017">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2369.700" EndTime="2370.600" Participant="mn017">
um -
</Segment>
<Segment StartTime="2371.942" EndTime="2376.890" Participant="mn017">
O_K. Then we have a messy alignment process where we actually insert into the sequence of words
</Segment>
<Segment StartTime="2377.499" EndTime="2383.987" Participant="mn017">
the, uh, <Emphasis> tags </Emphasis> for, like, where - where sentence - ends of sentence, question marks, um,
</Segment>
<Segment StartTime="2383.987" EndTime="2384.697" Participant="mn017">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2384.697" EndTime="2386.530" Participant="mn017">
various other things. Uh.
</Segment>
<Segment StartTime="2385.547" EndTime="2388.150" Participant="fe016">
Yeah. These are things that we had Don - So, Don sort of,
</Segment>
<Segment StartTime="2388.150" EndTime="2389.144" Participant="mn017">
Right. <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="2388.600" EndTime="2394.881" Participant="fe016">
um, propagated the punctuation from the original transcriber - so whether it was, like, question mark or period or,
</Segment>
<Segment StartTime="2394.881" EndTime="2395.379" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2395.379" EndTime="2400.206" Participant="fe016">
um, you know, comma and things like that, and we kept the - and disfluency dashes -
</Segment>
<Segment StartTime="2400.300" EndTime="2400.725" Participant="mn017">
Mm-hmm. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="2401.213" EndTime="2404.741" Participant="fe016">
uh, kept those in because we sort of wanna know where those are relative to the
</Segment>
<Segment StartTime="2402.813" EndTime="2403.802" Participant="mn017">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="2405.628" EndTime="2409.380" Participant="fe016">
spurt overlaps - sp- overlaps, or - <VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2406.533" EndTime="2410.582" Participant="mn017">
Right. So - so those are actually sort of retro-fitted into the time alignment.
</Segment>
<Segment StartTime="2407.408" EndTime="2410.044" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2408.858" EndTime="2409.669" Participant="me001">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2411.390" EndTime="2416.498" Participant="mn017">
And then we merge all the alignments from the various channels and we sort them by time.
</Segment>
<Segment StartTime="2417.716" EndTime="2418.515" Participant="mn017">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2418.515" EndTime="2419.862" Participant="mn017">
And then
</Segment>
<Segment StartTime="2420.340" EndTime="2421.139" Participant="mn017">
there's a -
</Segment>
<Segment StartTime="2421.691" EndTime="2425.486" Participant="mn017">
then there's a process where you now determine the spurts.
</Segment>
<Segment StartTime="2426.063" EndTime="2426.809" Participant="mn017">
That is -
</Segment>
<Segment StartTime="2426.809" EndTime="2433.037" Participant="mn017">
Actually, no, you do that before you merge the various channels. So you - you id- identify by some criterion, which is pause length -
</Segment>
<Segment StartTime="2433.037" EndTime="2439.448" Participant="mn017">
you identify the beginnings and ends of these spurts, and you put another set of tags in there to keep those straight.
</Segment>
<Segment StartTime="2437.790" EndTime="2438.421" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="2439.448" EndTime="2440.013" Participant="mn017">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2440.013" EndTime="2444.810" Participant="mn017">
And then you <Emphasis> merge </Emphasis> everything in terms of, you know, linearizing the sequence based on the <Emphasis> time </Emphasis> marks.
</Segment>
<Segment StartTime="2445.410" EndTime="2446.102" Participant="mn017">
And <Emphasis> then </Emphasis>
</Segment>
<Segment StartTime="2446.412" EndTime="2447.934" Participant="mn017">
<VocalSound Description="mouth"/> you extract
</Segment>
<Segment StartTime="2448.430" EndTime="2454.486" Participant="mn017">
the individual channels again, but this time you <Emphasis> know </Emphasis> where the other people start and end talking -
</Segment>
<Segment StartTime="2454.486" EndTime="2456.288" Participant="mn017">
you know, where their spurts start and end.
</Segment>
<Segment StartTime="2457.163" EndTime="2463.049" Participant="mn017">
And so you extract the individual <Emphasis> channels, </Emphasis> uh, one sp- spurt by spurt as it were. <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2462.752" EndTime="2463.919" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2463.049" EndTime="2469.259" Participant="mn017">
Um, and inside the words or between the words you now have begin and end <Pause/> tags for overlaps. So,
</Segment>
<Segment StartTime="2463.100" EndTime="2464.410" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2469.921" EndTime="2479.443" Participant="mn017">
you - you basically have everything sort of lined up and in a form where you can look at the individual speakers and how their speech relates to the other speakers' speech.
</Segment>
<Segment StartTime="2478.767" EndTime="2479.338" Participant="me011">
Right.
</Segment>
<Segment StartTime="2479.747" EndTime="2483.076" Participant="fe016">
Uh, I mean, I think that's actually really u- useful <Emphasis> also </Emphasis> because
</Segment>
<Segment StartTime="2479.880" EndTime="2480.509" Participant="mn017">
And -
</Segment>
<Segment StartTime="2483.784" EndTime="2490.133" Participant="fe016">
even if you weren't studying overlaps, if you wanna get a transcription for the far-field mikes, how are you gonna <Emphasis> know </Emphasis>
</Segment>
<Segment StartTime="2490.727" EndTime="2499.706" Participant="fe016">
which words from which speakers occurred at which times <Emphasis> relative </Emphasis> to each other? You <Emphasis> have </Emphasis> to be able to <Pause/> get a transcript like - like this anyway, just for doing far-field recognition. So,
</Segment>
<Segment StartTime="2500.389" EndTime="2501.056" Participant="fe016">
you know,
</Segment>
<Segment StartTime="2500.793" EndTime="2501.382" Participant="mn017">
Yeah.
</Segment>
<Segment StartTime="2501.056" EndTime="2501.993" Participant="fe016">
it's - it's sort of -
</Segment>
<Segment StartTime="2502.411" EndTime="2507.199" Participant="fe016">
I thi- it's just an issue we haven't <Emphasis> dealt </Emphasis> with before, how you <Emphasis> time-align </Emphasis> things that are overlapping anyway.
</Segment>
<Segment StartTime="2506.037" EndTime="2508.213" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2508.508" EndTime="2509.457" Participant="fe008">
That's wonderful.
</Segment>
<Segment StartTime="2508.540" EndTime="2509.057" Participant="mn017">
So -
</Segment>
<Segment StartTime="2508.540" EndTime="2510.660" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2509.107" EndTime="2509.640" Participant="me011">
Well -
</Segment>
<Segment StartTime="2509.457" EndTime="2510.292" Participant="mn017">
And - and we -
</Segment>
<Segment StartTime="2509.533" EndTime="2511.756" Participant="fe016">
I mean, i- I never thought about it before, but -
</Segment>
<Segment StartTime="2510.591" EndTime="2510.986" Participant="me011">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2510.719" EndTime="2511.945" Participant="me013">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2510.986" EndTime="2518.775" Participant="me011">
Y- yes. I mean, s- when I came up with the original data - suggested data format based on the transcription graph, there's capability of doing
</Segment>
<Segment StartTime="2511.204" EndTime="2511.569" Participant="mn017">
In -
</Segment>
<Segment StartTime="2511.928" EndTime="2512.918" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2512.035" EndTime="2514.607" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2516.446" EndTime="2517.180" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2517.752" EndTime="2518.560" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2519.205" EndTime="2520.690" Participant="me011">
<Emphasis> that </Emphasis> sort of thing <Emphasis> in </Emphasis> there.
</Segment>
<Segment StartTime="2520.022" EndTime="2528.525" Participant="fe016">
Right. But you <Emphasis> can't </Emphasis> get it directly from the <Emphasis> transcription. </Emphasis> Yeah, this is like a poor man's ver- <Emphasis> formatting </Emphasis> version. But it's, you know - It's <Emphasis> clean, </Emphasis> it's just not <Emphasis> fancy. </Emphasis>
</Segment>
<Segment StartTime="2520.226" EndTime="2534.305" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2520.629" EndTime="2521.520" Participant="fe008">
Mm-hmm. Yeah, that's right.
</Segment>
<Segment StartTime="2520.750" EndTime="2523.111" Participant="mn017">
Right. Well, this is - this is just -
</Segment>
<Segment StartTime="2528.370" EndTime="2536.085" Participant="mn017">
<VocalSound Description="inbreath"/> Well, there's lots of little things. It's like there're twelve different scripts which you run and then at the end you have what you want. But, um,
</Segment>
<Segment StartTime="2528.430" EndTime="2528.865" Participant="me011">
Right.
</Segment>
<Segment StartTime="2528.525" EndTime="2529.145" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2529.145" EndTime="2529.757" Participant="fe016">
Um.
</Segment>
<Segment StartTime="2532.736" EndTime="2534.386" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2533.807" EndTime="2534.183" Participant="fe016">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2535.985" EndTime="2539.102" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2536.603" EndTime="2541.551" Participant="mn017">
at the very last stage we throw away the actual <Emphasis> time </Emphasis> information. All we care about is whether -
</Segment>
<Segment StartTime="2542.015" EndTime="2545.873" Participant="mn017">
that there's a certain word was overlapped by someone <Emphasis> else's </Emphasis> word.
</Segment>
<Segment StartTime="2543.234" EndTime="2546.602" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2545.873" EndTime="2551.429" Participant="mn017">
So you sort of - at that point, you discretize things into just having overlap or no overlap.
</Segment>
<Segment StartTime="2551.214" EndTime="2551.864" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2552.007" EndTime="2556.804" Participant="mn017">
Because we figure that's about the level of analysis that we want to do for this paper. <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2555.717" EndTime="2556.350" Participant="me011">
Mm-hmm.
</Segment>
<Segment StartTime="2557.293" EndTime="2565.492" Participant="mn017">
But if you wanted to do a more <Emphasis> fine-grained </Emphasis> analysis and say, you know, how <Emphasis> far </Emphasis> into the word is the overlap, you could do that. It's just - it'll just require more -
</Segment>
<Segment StartTime="2563.595" EndTime="2567.540" Participant="fe016">
Yeah. Just <Pause/> sort of huge. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2566.190" EndTime="2567.490" Participant="mn017">
you know, slightly different -
</Segment>
<Segment StartTime="2567.190" EndTime="2574.127" Participant="fe008">
<VocalSound Description="breath"/> What's interesting is it's exactly what, um, i- in discussing with, um, <Emphasis> Sue </Emphasis> about this, um, she, um,
</Segment>
<Segment StartTime="2567.466" EndTime="2568.139" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2568.279" EndTime="2568.926" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="2574.127" EndTime="2575.717" Participant="fe008">
i- i- i- indicated that that -
</Segment>
<Segment StartTime="2576.210" EndTime="2578.486" Participant="fe008">
you know, that's very important for overlap analysis.
</Segment>
<Segment StartTime="2577.209" EndTime="2583.064" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2577.369" EndTime="2587.325" Participant="fe016">
Yeah. It's - it's nice to know, and also I think as a human, like, I don't always <Emphasis> hear </Emphasis> these in the actual order that they occur. So I can have two foreground speakers, you know, Morgan an- and
</Segment>
<Segment StartTime="2577.458" EndTime="2577.907" Participant="mn017">
Right.
</Segment>
<Segment StartTime="2587.325" EndTime="2587.882" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2587.882" EndTime="2597.061" Participant="fe016">
um, Adam and Jane could all be talking, and I could align each of them to be starting their utterance at the correct time, and then look where they are <Emphasis> relative </Emphasis> to each other, and that's not really what I <Emphasis> heard. </Emphasis>
</Segment>
<Segment StartTime="2597.090" EndTime="2601.941" Participant="fe008">
<VocalSound Description="inbreath"/> And that's <Emphasis> another </Emphasis> thing she said. This is - This is Bever's - Bever's effect, when - where -
</Segment>
<Segment StartTime="2598.427" EndTime="2599.910" Participant="fe016">
Cuz it's just <Emphasis> hard </Emphasis> to <Emphasis> do. </Emphasis>
</Segment>
<Segment StartTime="2601.301" EndTime="2602.610" Participant="fe016">
Y- Yeah. It's sort of -
</Segment>
<Segment StartTime="2602.850" EndTime="2608.646" Participant="fe008">
In psy- ps- psycho-linguistics you have these experiments where people have perceptual biases a- as to what they hear, that - that -
</Segment>
<Segment StartTime="2607.605" EndTime="2613.778" Participant="fe016">
Yeah, you sort of move things around until you get to a <Pause/> low information point and yo- then you can bring in the other person. So it's
</Segment>
<Segment StartTime="2609.364" EndTime="2610.229" Participant="fe008">
Not the best -
</Segment>
<Segment StartTime="2613.778" EndTime="2620.989" Participant="fe016">
<VocalSound Description="inbreath"/> actually not even <Emphasis> possible, </Emphasis> I think, for any <Emphasis> person </Emphasis> to listen to a mixed signal, even equalize, and make sure that they have all the words in the right order.
</Segment>
<Segment StartTime="2616.169" EndTime="2619.536" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2621.258" EndTime="2621.722" Participant="fe008">
Mm-hmm.
</Segment>
<Segment StartTime="2621.790" EndTime="2627.782" Participant="fe016">
So, I guess, we'll <Emphasis> try </Emphasis> to write this Eurospeech paper. I mean, we will <Emphasis> write </Emphasis> it. Whether they accept it <Pause/> late or not, I don't know.
</Segment>
<Segment StartTime="2622.276" EndTime="2623.069" Participant="fe008">
Superb.
</Segment>
<Segment StartTime="2626.155" EndTime="2626.570" Participant="me011">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2627.782" EndTime="2628.455" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2628.455" EndTime="2634.745" Participant="fe016">
Um, and the <Emphasis> good </Emphasis> thing is that we have - It's sort of a beginning of what Don can use to link the prosodic
</Segment>
<Segment StartTime="2635.219" EndTime="2637.079" Participant="fe016">
features from each file to each other.
</Segment>
<Segment StartTime="2635.904" EndTime="2636.542" Participant="mn017">
Yeah. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="2637.236" EndTime="2639.293" Participant="me013">
Yeah. That's the good thing about these pape-
</Segment>
<Segment StartTime="2637.981" EndTime="2638.565" Participant="mn005">
Hmm?
</Segment>
<Segment StartTime="2638.023" EndTime="2638.790" Participant="mn017">
Plus, mayb-
</Segment>
<Segment StartTime="2638.061" EndTime="2638.665" Participant="fe016">
So.
</Segment>
<Segment StartTime="2638.665" EndTime="2642.051" Participant="fe016">
i- You know, might as well. We- I ju- Otherwise we won't get the work done <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2640.262" EndTime="2640.856" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2640.271" EndTime="2642.460" Participant="mn005">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="2640.649" EndTime="2641.421" Participant="mn017">
I don't know, m-
</Segment>
<Segment StartTime="2641.915" EndTime="2648.537" Participant="mn017">
I mean, u- u- <Emphasis> Jane </Emphasis> likes to look at data. Maybe, you know, you could - you could look at this <Emphasis> format </Emphasis> and see if you find anything interesting.
</Segment>
<Segment StartTime="2642.051" EndTime="2643.837" Participant="fe016">
<VocalSound Description="laugh"/> on our deadline. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2642.319" EndTime="2642.972" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="2644.616" EndTime="2645.420" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2648.537" EndTime="2649.190" Participant="mn017">
I don't know.
</Segment>
<Segment StartTime="2648.929" EndTime="2657.616" Participant="me013">
Yeah. <VocalSound Description="inbreath"/> No, it's - that's the good thing about these pape- paper deadlines and, uh, you know, class projects, and - <VocalSound Description="laugh"/> and things like that, because you - you really get g- Yeah.
</Segment>
<Segment StartTime="2649.203" EndTime="2650.746" Participant="fe008">
<VocalSound Description="inbreath"/> Well, what I'm <Emphasis> thinking </Emphasis> is - <Comment Description="abandons to overlap"/>
</Segment>
<Segment StartTime="2649.861" EndTime="2650.483" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="2649.862" EndTime="2650.715" Participant="mn017">
Yeah.
</Segment>
<Segment StartTime="2650.983" EndTime="2651.497" Participant="fe008">
Yeah.
</Segment>
<Segment StartTime="2653.470" EndTime="2655.060" Participant="fe016">
Right. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2653.520" EndTime="2655.101" Participant="fe008">
<VocalSound Description="inbreath"/> Well, my - <Comment Description="abandons to overlap"/>
</Segment>
<Segment StartTime="2653.538" EndTime="2654.448" Participant="mn017">
Mm-hmm. <Comment Description="very quietly"/>
</Segment>
<Segment StartTime="2653.541" EndTime="2654.451" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2654.325" EndTime="2654.918" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2654.448" EndTime="2661.391" Participant="mn017">
Well th- th- the other thing that - that - that yo- that you usually don't tell your graduate students is that these deadlines are actually not that,
</Segment>
<Segment StartTime="2656.319" EndTime="2656.734" Participant="fe008">
Yeah.
</Segment>
<Segment StartTime="2656.358" EndTime="2658.750" Participant="fe016">
Forces you to do the work. <Comment Description="while laughing"/> Exactly.
</Segment>
<Segment StartTime="2661.391" EndTime="2664.931" Participant="mn017">
um, you know, strictly enforced, because <Pause/> the -
</Segment>
<Segment StartTime="2661.679" EndTime="2662.134" Participant="fe016">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2661.768" EndTime="2662.638" Participant="me011">
Strict.
</Segment>
<Segment StartTime="2663.696" EndTime="2664.230" Participant="fe008">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2663.762" EndTime="2667.592" Participant="me013">
<Uncertain NumSyllables="1"> @@ </Uncertain> <Comment Description="uncodeable, vocal squeak"/> <Emphasis> Oh, </Emphasis> <Emphasis> now </Emphasis> it's out in the <Emphasis> public, </Emphasis> this - this - this secret information. Yeah.
</Segment>
<Segment StartTime="2665.277" EndTime="2668.503" Participant="fe008">
<VocalSound Description="laugh"/> I think we can ha- <Comment Description="quietly"/>
</Segment>
<Segment StartTime="2665.417" EndTime="2669.480" Participant="fe016">
<VocalSound Description="laugh"/> Right. So - <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2665.468" EndTime="2666.584" Participant="mn017">
because - <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2665.650" EndTime="2670.445" Participant="me026">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2665.860" EndTime="2666.629" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2666.584" EndTime="2670.156" Participant="mn017">
<VocalSound Description="laugh"/> bec- b- <VocalSound Description="laugh"/> <Uncertain> Nah </Uncertain> - <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2667.527" EndTime="2670.534" Participant="me011">
<VocalSound Description="laugh"/> No. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2667.592" EndTime="2672.480" Participant="me013">
<VocalSound Description="laugh"/> No. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2668.503" EndTime="2671.377" Participant="fe008">
<VocalSound Description="breath-laugh"/> Nah. <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2668.970" EndTime="2671.580" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2669.569" EndTime="2671.455" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2670.393" EndTime="2676.571" Participant="mn017">
i- Because these - the <Emphasis> conference </Emphasis> organizers actually have an interest in getting lots of submissions. I mean, a - a <Emphasis> monetary </Emphasis> interest. So -
</Segment>
<Segment StartTime="2670.819" EndTime="2672.370" Participant="fe016">
Right. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2672.846" EndTime="2676.108" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2674.689" EndTime="2675.202" Participant="me011">
Right.
</Segment>
<Segment StartTime="2675.578" EndTime="2676.072" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="2676.571" EndTime="2677.287" Participant="mn017">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2677.287" EndTime="2678.304" Participant="mn017">
Um. <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="2677.852" EndTime="2679.050" Participant="me013">
Th- that's - that's true.
</Segment>
<Segment StartTime="2677.993" EndTime="2683.167" Participant="fe008">
And <Emphasis> good </Emphasis> ones, <Emphasis> good </Emphasis> ones, which sometimes means <Pause/> a little extra time. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2679.050" EndTime="2680.759" Participant="mn017">
And <Emphasis> good </Emphasis> submission- Right. Well -
</Segment>
<Segment StartTime="2679.474" EndTime="2680.747" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2679.494" EndTime="2679.929" Participant="me013">
That's -
</Segment>
<Segment StartTime="2680.602" EndTime="2681.175" Participant="me013">
That's true.
</Segment>
<Segment StartTime="2680.759" EndTime="2682.820" Participant="mn017">
<VocalSound Description="inbreath"/> That's another issue, but -
</Segment>
<Segment StartTime="2681.413" EndTime="2682.204" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2682.204" EndTime="2689.660" Participant="me013">
By th- by the way, this is <Emphasis> totally </Emphasis> unfair, you may - you may feel, but the - the, uh - the <Emphasis> morning </Emphasis> meeting folks actually have an - an extra month or so.
</Segment>
<Segment StartTime="2685.661" EndTime="2686.116" Participant="mn017">
<Uncertain> Mm-hmm. </Uncertain>
</Segment>
<Segment StartTime="2689.580" EndTime="2692.863" Participant="mn014">
Yep. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2690.752" EndTime="2693.738" Participant="me011">
Yep. The Aurora - there's a special Aurora -
</Segment>
<Segment StartTime="2691.536" EndTime="2691.951" Participant="mn017">
When -
</Segment>
<Segment StartTime="2691.606" EndTime="2692.160" Participant="fe016">
Uh -
</Segment>
<Segment StartTime="2692.761" EndTime="2694.106" Participant="me026">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2693.172" EndTime="2700.140" Participant="me013">
There's a special Aurora session and the Aurora pe- people involved in Aurora have till Ma- uh, early May <Pause/> or something to turn in their paper.
</Segment>
<Segment StartTime="2694.770" EndTime="2695.308" Participant="fe016">
Oh.
</Segment>
<Segment StartTime="2694.773" EndTime="2695.646" Participant="mn017">
<Emphasis> Mmm. </Emphasis>
</Segment>
<Segment StartTime="2695.334" EndTime="2697.122" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2695.929" EndTime="2697.490" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2695.992" EndTime="2696.624" Participant="fe016">
Oh.
</Segment>
<Segment StartTime="2699.006" EndTime="2699.576" Participant="mn017">
Mmm.
</Segment>
<Segment StartTime="2700.042" EndTime="2704.016" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2700.292" EndTime="2701.162" Participant="mn017">
Well, then you can just -
</Segment>
<Segment StartTime="2700.561" EndTime="2704.170" Participant="fe016">
Oh, well maybe we'll submit to s- <Comment Description="while laughing"/> <VocalSound Description="laugh"/> Actually -
</Segment>
<Segment StartTime="2700.656" EndTime="2703.742" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2701.025" EndTime="2704.915" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2701.298" EndTime="2704.233" Participant="me026">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2701.597" EndTime="2705.635" Participant="mn017">
Maybe you can submit the digits paper on e- for the Aurora session.
</Segment>
<Segment StartTime="2702.083" EndTime="2703.906" Participant="mn005">
Yeah. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2703.668" EndTime="2703.905" Participant="me011">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2703.742" EndTime="2704.452" Participant="mn014">
Yeah. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2704.170" EndTime="2704.800" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="2704.696" EndTime="2707.912" Participant="me011">
Oh, I could! I could submit that to Aurora. That would be pretty - pretty -
</Segment>
<Segment StartTime="2704.915" EndTime="2706.051" Participant="me013">
I- if it w-
</Segment>
<Segment StartTime="2705.290" EndTime="2706.016" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="2706.051" EndTime="2707.440" Participant="me013">
<VocalSound Description="inbreath"/> Well -
</Segment>
<Segment StartTime="2706.502" EndTime="2707.096" Participant="mn017">
<Uncertain> Yeah. </Uncertain>
</Segment>
<Segment StartTime="2707.440" EndTime="2708.704" Participant="me013">
i- it has -
</Segment>
<Segment StartTime="2708.272" EndTime="2709.081" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="2708.288" EndTime="2710.422" Participant="me011">
S- That wouldn't work. It's not Aurora.
</Segment>
<Segment StartTime="2708.704" EndTime="2709.260" Participant="me013">
<Uncertain NumSyllables="2"> @@ </Uncertain> <Comment Description="uncodeable, vocal squeaks"/>
</Segment>
<Segment StartTime="2709.260" EndTime="2713.657" Participant="me013">
No, it wouldn't work. It's - it's not the Aurora - I mean, it - it's - it's actually the Aurora <Emphasis> task. </Emphasis>
</Segment>
<Segment StartTime="2710.338" EndTime="2711.360" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2711.833" EndTime="2712.980" Participant="fe016">
Maybe they'll get s-
</Segment>
<Segment StartTime="2712.103" EndTime="2713.922" Participant="me011">
Aurora's very specific.
</Segment>
<Segment StartTime="2713.657" EndTime="2713.958" Participant="me013">
It-
</Segment>
<Segment StartTime="2713.692" EndTime="2718.307" Participant="fe016">
Well, maybe it won't be after this <VocalSound Description="laugh"/> deadline <Pause/> extension. Maybe they'll - <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2714.874" EndTime="2716.399" Participant="mn017">
But - but the people - I mean,
</Segment>
<Segment StartTime="2716.380" EndTime="2717.895" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2716.399" EndTime="2723.507" Participant="mn017">
a - a paper that is <Emphasis> not </Emphasis> on Aurora would probably be <Emphasis> more </Emphasis> interesting at that point because everybody's so sick and tired of the Aurora task.
</Segment>
<Segment StartTime="2722.259" EndTime="2723.323" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2722.463" EndTime="2727.268" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2723.249" EndTime="2723.724" Participant="mn014">
Yeah. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="2723.323" EndTime="2728.580" Participant="me011">
<VocalSound Description="inbreath"/> Oh, I thought you meant this was just the <Emphasis> digits </Emphasis> section. I didn't know you meant it was <Emphasis> Aurora </Emphasis> digits. <VocalSound Description="breath"/>
</Segment>
<Segment StartTime="2723.329" EndTime="2724.039" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2723.793" EndTime="2728.005" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2727.759" EndTime="2729.816" Participant="mn014">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2728.005" EndTime="2728.604" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="2728.129" EndTime="2728.932" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2728.551" EndTime="2730.203" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2728.580" EndTime="2733.697" Participant="mn017">
Well, no. If you - if you have - it's to - if you discuss some relation to the Aurora task, like if you use the same -
</Segment>
<Segment StartTime="2728.941" EndTime="2729.574" Participant="me011">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2730.180" EndTime="2730.754" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2733.118" EndTime="2735.017" Participant="me013">
<Emphasis> This </Emphasis> is <Emphasis> not </Emphasis> the Aurora task.
</Segment>
<Segment StartTime="2734.736" EndTime="2735.409" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2735.017" EndTime="2738.292" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2735.082" EndTime="2742.271" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2735.177" EndTime="2739.412" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2735.208" EndTime="2736.395" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2735.275" EndTime="2740.911" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2736.565" EndTime="2738.940" Participant="me026">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2736.692" EndTime="2739.085" Participant="mn017">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2738.292" EndTime="2746.630" Participant="me013">
So they just do a little grep for - <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2739.412" EndTime="2740.282" Participant="fe016">
Do - uh, d- d-
</Segment>
<Segment StartTime="2740.114" EndTime="2741.110" Participant="mn017">
Um.
</Segment>
<Segment StartTime="2740.282" EndTime="2745.484" Participant="fe016">
Do not - do not - we are not setting a good example. This is not a - <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2741.840" EndTime="2744.670" Participant="mn017">
Well, a relation other than negation, maybe, um.
</Segment>
<Segment StartTime="2744.173" EndTime="2751.764" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2744.405" EndTime="2745.869" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2745.310" EndTime="2746.766" Participant="mn017">
So. I don't know.
</Segment>
<Segment StartTime="2746.117" EndTime="2746.750" Participant="fe016">
Anyway.
</Segment>
<Segment StartTime="2746.630" EndTime="2747.950" Participant="me013">
<VocalSound Description="clears throat"/>
</Segment>
<Segment StartTime="2747.383" EndTime="2749.302" Participant="fe016">
But the good thing is this does -
</Segment>
<Segment StartTime="2747.572" EndTime="2750.077" Participant="me026">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2747.651" EndTime="2754.683" Participant="me011">
<VocalSound Description="inbreath"/> Well, I- I don't know. I mean, you could - you could do a paper on <Pause/> what's wrong with the Aurora task by comparing it to <Pause/> other ways of doing it. <VocalSound Description="breath"/>
</Segment>
<Segment StartTime="2750.043" EndTime="2750.379" Participant="fe008">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2753.398" EndTime="2754.139" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2753.602" EndTime="2753.940" Participant="mn017">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2753.940" EndTime="2761.488" Participant="mn017">
How well does an Aurora system do on - on - you know, on digits collected in a - in this environment? Yeah.
</Segment>
<Segment StartTime="2754.679" EndTime="2755.896" Participant="mn005">
<Uncertain NumSyllables="3"> @@ </Uncertain> <Comment Description="during mike spike"/>
</Segment>
<Segment StartTime="2755.211" EndTime="2758.030" Participant="me026">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="2755.228" EndTime="2755.544" Participant="fe016">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2756.178" EndTime="2758.205" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2759.989" EndTime="2761.017" Participant="me011">
Different way. Yeah.
</Segment>
<Segment StartTime="2761.658" EndTime="2762.293" Participant="me013">
Maybe.
</Segment>
<Segment StartTime="2762.293" EndTime="2764.390" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2762.377" EndTime="2763.007" Participant="mn017">
Maybe.
</Segment>
<Segment StartTime="2762.976" EndTime="2763.510" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2763.007" EndTime="2763.634" Participant="mn017">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2763.120" EndTime="2765.037" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2763.200" EndTime="2763.912" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2763.499" EndTime="2765.410" Participant="me011">
Pretty hokey. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2763.737" EndTime="2764.790" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2764.390" EndTime="2770.403" Participant="me013">
I think it's a littl- little far-fetched. <Comment Description="while laughing"/> Nah, I mean, the thing is Aurora's pretty closed community. I mean, you know, the people who were involved in the -
</Segment>
<Segment StartTime="2764.548" EndTime="2765.022" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2765.237" EndTime="2767.730" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2765.572" EndTime="2766.245" Participant="mn017">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2766.536" EndTime="2767.054" Participant="me011">
Yep.
</Segment>
<Segment StartTime="2768.727" EndTime="2769.710" Participant="mn017">
<NonVocalSound Description="mike noise"/> Mm-hmm.
</Segment>
<Segment StartTime="2770.403" EndTime="2770.897" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2770.897" EndTime="2776.703" Participant="me013">
the only people who are allowed to test on that are people who - who made it above a certain threshold in the first round,
</Segment>
<Segment StartTime="2771.159" EndTime="2772.544" Participant="me011">
It's very specific.
</Segment>
<Segment StartTime="2776.703" EndTime="2780.551" Participant="me013">
<VocalSound Description="inbreath"/> uh <VocalSound Description="mouth"/> <Uncertain> w- </Uncertain> in ninety-nine and it's - it's sort of a - it's -
</Segment>
<Segment StartTime="2780.551" EndTime="2782.114" Participant="me013">
<VocalSound Description="inbreath"/> not like a -
</Segment>
<Segment StartTime="2781.001" EndTime="2782.604" Participant="mn017">
Well, that's maybe why they don't f-
</Segment>
<Segment StartTime="2782.950" EndTime="2787.524" Participant="mn017">
know that they have a crummy system. I mean, a crummy back-end. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2784.178" EndTime="2791.720" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2784.298" EndTime="2791.647" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2784.491" EndTime="2786.363" Participant="fe008">
<VocalSound Description="mock gasp"/> <Comment Description="vocal gesture of mock horror, regarding bleep-potential of [mn017]'s utterance"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2785.983" EndTime="2786.853" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2787.524" EndTime="2790.963" Participant="mn017">
No, I mean - I mean, <Emphasis> seriously, </Emphasis> if you - if you have a very - <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2788.776" EndTime="2791.323" Participant="me026">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2789.154" EndTime="2790.969" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2789.871" EndTime="2792.170" Participant="fe008">
<VocalSound Description="inbreath"/> <VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2790.963" EndTime="2791.892" Participant="mn017">
No, I'm sorry.
</Segment>
<Segment StartTime="2791.333" EndTime="2793.016" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2791.647" EndTime="2794.902" Participant="fe016">
Uh, <Comment Description="while laughing"/> "beep" <VocalSound Description="laugh"/> "bee-" <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2791.720" EndTime="2795.580" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2791.892" EndTime="2795.370" Participant="mn017">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2792.777" EndTime="2794.990" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2792.843" EndTime="2794.168" Participant="fe008">
<VocalSound Description="laugh"/> <Comment Description="concurrent mike noise"/>
</Segment>
<Segment StartTime="2793.293" EndTime="2795.982" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2793.341" EndTime="2795.101" Participant="me026">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2794.902" EndTime="2798.849" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2795.370" EndTime="2796.420" Participant="mn017">
No. I didn't mean
</Segment>
<Segment StartTime="2795.790" EndTime="2796.858" Participant="me011">
I mean, th- <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="2796.311" EndTime="2797.446" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2796.857" EndTime="2800.631" Participant="mn017">
anybody - any <Emphasis> particular </Emphasis> system. I meant this H_T_<Emphasis> K </Emphasis> back-end. If they -
</Segment>
<Segment StartTime="2796.875" EndTime="2798.148" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2797.249" EndTime="2797.526" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2800.433" EndTime="2804.013" Participant="me013">
Oh, you don't like H_T_K? <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2800.811" EndTime="2801.781" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2801.105" EndTime="2806.965" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2801.113" EndTime="2804.780" Participant="me026">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2801.679" EndTime="2804.335" Participant="mn005">
Yeah. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2801.705" EndTime="2803.980" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2801.785" EndTime="2803.488" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2801.877" EndTime="2803.222" Participant="mn017">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2801.982" EndTime="2804.080" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2803.542" EndTime="2807.257" Participant="mn017">
I don't h- I don't have any stock in H_T_K or Entropic or anything.
</Segment>
<Segment StartTime="2803.700" EndTime="2805.381" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2804.784" EndTime="2812.340" Participant="me013">
<VocalSound Description="laugh"/> No. I mean, this - it- it's the H_T_K <Pause/> that is trained on a very limited amount of data. Yeah.
</Segment>
<Segment StartTime="2805.624" EndTime="2807.543" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2807.257" EndTime="2807.970" Participant="mn017">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2807.945" EndTime="2809.648" Participant="me011">
It's d- it's very specific.
</Segment>
<Segment StartTime="2809.060" EndTime="2809.851" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2810.411" EndTime="2811.479" Participant="fe008">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2811.150" EndTime="2811.461" Participant="mn017">
Right.
</Segment>
<Segment StartTime="2812.312" EndTime="2813.448" Participant="mn017">
But so, if you -
</Segment>
<Segment StartTime="2814.072" EndTime="2818.505" Participant="mn017">
But maybe you should, you know, consider <Emphasis> more </Emphasis> - using <Emphasis> more </Emphasis> data, or - I mean -
</Segment>
<Segment StartTime="2814.268" EndTime="2814.703" Participant="fe008">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2817.679" EndTime="2819.809" Participant="me013">
Oh, yeah. I - I really think that that's <Emphasis> true. </Emphasis>
</Segment>
<Segment StartTime="2819.809" EndTime="2821.089" Participant="me013">
And they i- i- <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="2819.810" EndTime="2824.835" Participant="mn017">
If yo- if you sort of hermetically stay within one task and don't look left and right, then you're gonna -
</Segment>
<Segment StartTime="2820.468" EndTime="2821.516" Participant="me011">
But they - they had -
</Segment>
<Segment StartTime="2821.089" EndTime="2822.499" Participant="me013">
<VocalSound Description="inbreath"/> <VocalSound Description="inbreath"/> i- <Comment Description="begins to speak"/>
</Segment>
<Segment StartTime="2822.912" EndTime="2823.313" Participant="me013">
But -
</Segment>
<Segment StartTime="2823.331" EndTime="2826.128" Participant="me011">
They had something very specific in mind when they designed it.
</Segment>
<Segment StartTime="2826.117" EndTime="2826.512" Participant="mn017">
Right.
</Segment>
<Segment StartTime="2826.534" EndTime="2827.421" Participant="me013">
Well, u- i-
</Segment>
<Segment StartTime="2826.674" EndTime="2831.169" Participant="me011">
Right? And so - so you can - you can argue about maybe that wasn't the right thing to <Emphasis> do, </Emphasis> but, you know, they - they -
</Segment>
<Segment StartTime="2829.546" EndTime="2831.702" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2830.663" EndTime="2831.118" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2831.420" EndTime="2836.540" Participant="me013">
But, one of the reasons I have Chuck's messing around with - with the back-end that you're not supposed to touch -
</Segment>
<Segment StartTime="2831.738" EndTime="2833.300" Participant="me011">
they had something specific.
</Segment>
<Segment StartTime="2836.453" EndTime="2837.149" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="2836.540" EndTime="2840.052" Participant="me013">
I mean, for the evaluations, yes, we'll run a version that hasn't been touched.
</Segment>
<Segment StartTime="2838.001" EndTime="2838.735" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="2840.052" EndTime="2840.479" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2840.479" EndTime="2846.799" Participant="me013">
But, uh, one of the reasons I <Emphasis> have </Emphasis> him messing around with that, because I think it's sort of an open question that we <Emphasis> don't </Emphasis> know the answer to. People always say very glibly
</Segment>
<Segment StartTime="2840.761" EndTime="2841.493" Participant="mn017">
<VocalSound Description="clears throat"/>
</Segment>
<Segment StartTime="2841.048" EndTime="2844.510" Participant="mn005">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="2846.799" EndTime="2847.135" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2847.135" EndTime="2851.605" Participant="me013">
that i- if you s- show improvement on a bad system, that doesn't <Emphasis> mean </Emphasis> anything, cuz it may not be -
</Segment>
<Segment StartTime="2851.605" EndTime="2855.006" Participant="me013">
<VocalSound Description="inbreath"/> show - uh, because, you know, it doesn't <Emphasis> tell </Emphasis> you anything about the <Emphasis> good </Emphasis> system.
</Segment>
<Segment StartTime="2854.476" EndTime="2855.226" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="2855.006" EndTime="2858.029" Participant="me013">
<VocalSound Description="inbreath"/> And I - <Emphasis> I've </Emphasis> always sort of felt that that <Emphasis> depends. </Emphasis>
</Segment>
<Segment StartTime="2858.490" EndTime="2863.986" Participant="me013">
You know, that if some peopl- If you're actually are getting at something that has some <Pause/> conceptual <Emphasis> substance </Emphasis> to it, it <Emphasis> will </Emphasis> port.
</Segment>
<Segment StartTime="2858.574" EndTime="2859.275" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="2861.676" EndTime="2867.393" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2864.286" EndTime="2870.808" Participant="me013">
And in fact, <Emphasis> most </Emphasis> methods that people now use <Emphasis> were </Emphasis> originally tried with something that was not their absolute <Pause/> best system at <Emphasis> some </Emphasis> level.
</Segment>
<Segment StartTime="2870.808" EndTime="2872.592" Participant="me013">
<VocalSound Description="inbreath"/> But of course, sometimes it doesn't,
</Segment>
<Segment StartTime="2873.007" EndTime="2878.221" Participant="me013">
uh, port. So I think that's - that's an interesting question. If we're getting <Pause/> three percent error on, uh,
</Segment>
<Segment StartTime="2878.221" EndTime="2881.315" Participant="me013">
u- uh, English, uh, nati- native speakers,
</Segment>
<Segment StartTime="2881.315" EndTime="2881.849" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2881.849" EndTime="2888.010" Participant="me013">
um, using the Aurora system, and we do some <Emphasis> improvements </Emphasis> and bring it from three to <Emphasis> two, </Emphasis>
</Segment>
<Segment StartTime="2888.010" EndTime="2888.425" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2888.425" EndTime="2893.669" Participant="me013">
do those same improvements bring, uh, th- you know, the S_R_I system from one point three to - you know, to
</Segment>
<Segment StartTime="2890.375" EndTime="2891.600" Participant="mn017">
Hmm. <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="2893.459" EndTime="2894.033" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="2893.669" EndTime="2894.566" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2894.050" EndTime="2894.586" Participant="me011">
Zero.
</Segment>
<Segment StartTime="2894.566" EndTime="2899.000" Participant="me013">
point eight? Well. <VocalSound Description="laugh"/> You know, so that's - that's something we can <Emphasis> test. </Emphasis> <VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2894.586" EndTime="2898.028" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2895.700" EndTime="2898.176" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2896.201" EndTime="2896.724" Participant="mn017">
Mmm.
</Segment>
<Segment StartTime="2898.186" EndTime="2898.799" Participant="mn017">
Right.
</Segment>
<Segment StartTime="2899.704" EndTime="2900.633" Participant="me013">
<VocalSound Description="mouth"/> So.
</Segment>
<Segment StartTime="2900.633" EndTime="2901.209" Participant="me013">
Anyway.
</Segment>
<Segment StartTime="2900.995" EndTime="2901.885" Participant="mn017">
O_K. <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="2901.126" EndTime="2901.963" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2901.209" EndTime="2908.440" Participant="me013">
<VocalSound Description="laugh"/> I think we've - <VocalSound Description="laugh"/> we've covered that one up extremely well. <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2901.758" EndTime="2902.173" Participant="fe008">
Mm-hmm.
</Segment>
<Segment StartTime="2903.670" EndTime="2906.057" Participant="mn017">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2903.902" EndTime="2904.387" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2905.056" EndTime="2909.317" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2905.139" EndTime="2905.815" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2906.057" EndTime="2906.492" Participant="mn017">
Whew! <Comment Description="vocal gesture of relief"/> <Comment Description="humor"/>
</Segment>
<Segment StartTime="2906.492" EndTime="2909.530" Participant="mn017">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2907.039" EndTime="2908.819" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2907.075" EndTime="2907.560" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2908.440" EndTime="2909.592" Participant="me013">
O_K. <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2909.592" EndTime="2911.409" Participant="me013">
So, um -
</Segment>
<Segment StartTime="2910.605" EndTime="2912.160" Participant="mn005">
<NonVocalSound Description="mike noise"/> <VocalSound Description="cough"/>
</Segment>
<Segment StartTime="2911.890" EndTime="2912.404" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2912.404" EndTime="2919.309" Participant="me013">
Yeah. So tha- so we'll - you know, maybe you guys'll have - have one. Uh, you - you and, uh - and Dan have - have a paper that -
</Segment>
<Segment StartTime="2919.309" EndTime="2924.117" Participant="me013">
that's going in. You know, that's - that's pretty solid, on the segmentation <Pause/> stuff.
</Segment>
<Segment StartTime="2919.404" EndTime="2920.751" Participant="mn014">
Yeah. Yeah.
</Segment>
<Segment StartTime="2922.698" EndTime="2926.583" Participant="mn014">
Yeah. I will send you the - the final version, which is not -
</Segment>
<Segment StartTime="2924.748" EndTime="2925.277" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="2925.717" EndTime="2929.930" Participant="me013">
And the Aurora folks here will - will definitely get something in on Aurora, so.
</Segment>
<Segment StartTime="2927.969" EndTime="2933.844" Participant="mn017">
<VocalSound Description="mouth"/> Actually this - this, um - So, there's another paper. It's a Eurospeech paper but not related to meetings.
</Segment>
<Segment StartTime="2934.200" EndTime="2935.714" Participant="mn017">
But it's on <Emphasis> digits. </Emphasis> So,
</Segment>
<Segment StartTime="2936.505" EndTime="2937.059" Participant="mn017">
um,
</Segment>
<Segment StartTime="2937.460" EndTime="2942.143" Participant="mn017">
uh, a colleague at S_R_I developed a improved version of M_M_I_E training.
</Segment>
<Segment StartTime="2942.297" EndTime="2942.810" Participant="me013">
Uh-huh.
</Segment>
<Segment StartTime="2942.810" EndTime="2949.181" Participant="mn017">
And he tested it mostly on digits because it's sort of a - you know, it doesn't take weeks to train it. <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2949.039" EndTime="2949.949" Participant="me013">
Right. <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="2949.181" EndTime="2949.933" Participant="mn017">
Um. <VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2949.933" EndTime="2953.295" Participant="mn017">
And got some very impressive results, um,
</Segment>
<Segment StartTime="2953.651" EndTime="2956.536" Participant="mn017">
with, you know, discriminative, uh, Gaussian training.
</Segment>
<Segment StartTime="2957.821" EndTime="2959.587" Participant="mn017">
Um, you know, like, um,
</Segment>
<Segment StartTime="2960.621" EndTime="2965.175" Participant="mn017">
error rates <Pause/> go from - I don't know, in very <Emphasis> noisy </Emphasis> environment, like from,
</Segment>
<Segment StartTime="2965.175" EndTime="2966.278" Participant="mn017">
uh,
</Segment>
<Segment StartTime="2966.842" EndTime="2967.238" Participant="me011">
<VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="2967.148" EndTime="2967.880" Participant="mn017">
uh - <Comment Description="quietly"/>
</Segment>
<Segment StartTime="2970.109" EndTime="2975.299" Participant="mn017">
I for- now I - O_K, now I have the order of magnit- I'm not sure about the order of magnitude. Was it like from ten percent to
</Segment>
<Segment StartTime="2971.289" EndTime="2971.645" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="2975.299" EndTime="2975.794" Participant="mn017">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2975.794" EndTime="2977.201" Participant="mn017">
eight percent or from
</Segment>
<Segment StartTime="2977.201" EndTime="2978.484" Participant="mn017">
e- e- you know, point -
</Segment>
<Segment StartTime="2978.840" EndTime="2981.933" Participant="mn017">
you know, from one percent to point eight percent? I mean, it's a - <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2978.989" EndTime="2980.217" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2980.217" EndTime="2984.707" Participant="me013">
H- i- it got - it got better. Yeah, yeah. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2980.719" EndTime="2984.489" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2981.274" EndTime="2983.360" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2982.190" EndTime="2984.920" Participant="mn017">
It got <Emphasis> better. </Emphasis> That's the important thing. Yeah. But it's -
</Segment>
<Segment StartTime="2982.262" EndTime="2987.108" Participant="me011">
<VocalSound Description="laugh"/> Hey, that's the same percent relative, so - <VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2982.472" EndTime="2983.144" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="2984.707" EndTime="2986.369" Participant="me013">
<VocalSound Description="laugh"/> Yeah.
</Segment>
<Segment StartTime="2985.335" EndTime="2988.065" Participant="mn017">
Yeah. Right. It's, uh, something in - Right.
</Segment>
<Segment StartTime="2985.714" EndTime="2986.347" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2986.369" EndTime="2987.690" Participant="me013">
<VocalSound Description="breath-laugh"/> Yeah.
</Segment>
<Segment StartTime="2987.602" EndTime="2989.246" Participant="me011">
Twenty percent relative gain.
</Segment>
<Segment StartTime="2988.808" EndTime="2989.329" Participant="mn017">
Yeah.
</Segment>
<Segment StartTime="2989.151" EndTime="2990.558" Participant="me013">
Yeah. Yeah.
</Segment>
<Segment StartTime="2990.558" EndTime="2991.927" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2991.927" EndTime="2996.023" Participant="me013">
Um, <VocalSound Description="outbreath"/> let's see. I think the only thing we had left was - unless somebody else - Well,
</Segment>
<Segment StartTime="2996.023" EndTime="3000.177" Participant="me013">
there's a <Emphasis> couple </Emphasis> things. Uh, one is <Pause/> anything that, um,
</Segment>
<Segment StartTime="3000.711" EndTime="3004.370" Participant="me013">
<VocalSound Description="mouth"/> anybody has to say about Saturday? Anything we should do in prep for Saturday?
</Segment>
<Segment StartTime="3005.161" EndTime="3006.220" Participant="me013">
Um -
</Segment>
<Segment StartTime="3006.220" EndTime="3007.138" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3007.138" EndTime="3008.426" Participant="me013">
I guess everybody knows about -
</Segment>
<Segment StartTime="3008.426" EndTime="3015.487" Participant="me013">
<VocalSound Description="inbreath"/> I mean, u- um, Mari was asking - was trying to come up with something like an agenda and we're sort of fitting around people's times a bit.
</Segment>
<Segment StartTime="3015.487" EndTime="3016.194" Participant="me013">
But, um,
</Segment>
<Segment StartTime="3016.194" EndTime="3016.812" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3016.812" EndTime="3021.227" Participant="me013">
clearly when we actually get here we'll <Pause/> move things around this, as we need to, but -
</Segment>
<Segment StartTime="3021.227" EndTime="3024.787" Participant="me013">
<VocalSound Description="mouth"/> so you can't <Emphasis> absolutely </Emphasis> count on it. But - but, uh - <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="3021.402" EndTime="3021.987" Participant="mn014">
O_K.
</Segment>
<Segment StartTime="3023.178" EndTime="3023.613" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="3024.492" EndTime="3026.730" Participant="fe016">
Are we meeting in <Emphasis> here </Emphasis> <Uncertain> probably </Uncertain> or - ? O_K.
</Segment>
<Segment StartTime="3025.365" EndTime="3026.374" Participant="me013">
<VocalSound Description="inbreath"/> Yeah.
</Segment>
<Segment StartTime="3026.728" EndTime="3028.943" Participant="me013">
That was my thought. I think this is -
</Segment>
<Segment StartTime="3027.809" EndTime="3028.877" Participant="mn017">
Are we recording it?
</Segment>
<Segment StartTime="3027.871" EndTime="3028.331" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="3028.943" EndTime="3029.813" Participant="me013">
<VocalSound Description="inbreath"/> <Uncertain NumSyllables="1"> @@ </Uncertain> <Comment Description="uncodeable, vocal squeak"/>
</Segment>
<Segment StartTime="3029.152" EndTime="3031.207" Participant="fe016">
We won't have enough microphones, but -
</Segment>
<Segment StartTime="3030.446" EndTime="3034.006" Participant="me013">
u- No. I - I <Emphasis> hadn't </Emphasis> in- intended to. We won- we wanna - I mean, they're -
</Segment>
<Segment StartTime="3032.079" EndTime="3032.749" Participant="fe016">
There's no way.
</Segment>
<Segment StartTime="3032.104" EndTime="3033.227" Participant="mn017">
O_K. <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="3034.006" EndTime="3038.615" Participant="me013">
there's gonna be, uh, Jeff, Katrin, Mari and two students. So there's five <Pause/> from there.
</Segment>
<Segment StartTime="3037.867" EndTime="3038.520" Participant="me011">
And Brian.
</Segment>
<Segment StartTime="3038.615" EndTime="3039.189" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3038.970" EndTime="3040.030" Participant="mn017">
But you know th- <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3039.189" EndTime="3041.450" Participant="me013">
And Brian's coming, so that's <Emphasis> six. </Emphasis> <Comment Description="with continuing list intonation"/>
</Segment>
<Segment StartTime="3040.734" EndTime="3042.178" Participant="mn017">
Mm-hmm. <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="3041.170" EndTime="3042.280" Participant="me011">
And plus all of us.
</Segment>
<Segment StartTime="3042.178" EndTime="3043.650" Participant="mn017">
Can use the Oprah mike.
</Segment>
<Segment StartTime="3042.215" EndTime="3044.100" Participant="me013">
Uh -
</Segment>
<Segment StartTime="3044.257" EndTime="3044.821" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3045.605" EndTime="3048.763" Participant="fe016">
Depends how fast you can <Pause/> throw it. It's just -
</Segment>
<Segment StartTime="3046.600" EndTime="3049.520" Participant="me011">
<VocalSound Description="inbreath"/> It seems like too many - too much coming and going.
</Segment>
<Segment StartTime="3048.952" EndTime="3049.619" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="3049.336" EndTime="3049.845" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="3050.269" EndTime="3050.664" Participant="me013">
Well -
</Segment>
<Segment StartTime="3050.444" EndTime="3052.165" Participant="fe016">
We don't even have enough <Emphasis> channel </Emphasis> -
</Segment>
<Segment StartTime="3051.054" EndTime="3053.606" Participant="mn017">
Because it would be a different kind of meeting, that's what I'm -
</Segment>
<Segment StartTime="3052.508" EndTime="3052.983" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="3053.445" EndTime="3054.364" Participant="me013">
<VocalSound Description="inbreath"/> Well -
</Segment>
<Segment StartTime="3053.731" EndTime="3055.079" Participant="mn005">
Yeah. <VocalSound Description="breath"/>
</Segment>
<Segment StartTime="3054.235" EndTime="3054.611" Participant="mn017">
But -
</Segment>
<Segment StartTime="3054.851" EndTime="3057.287" Participant="me013">
I hadn't <Pause/> really thought of it, but -
</Segment>
<Segment StartTime="3055.897" EndTime="3060.480" Participant="mn017">
<VocalSound Description="mouth"/> Maybe just - maybe not the whole day but just, you know, maybe some - I mean, part of it?
</Segment>
<Segment StartTime="3058.355" EndTime="3059.403" Participant="me013">
Maybe <Emphasis> part </Emphasis> of it.
</Segment>
<Segment StartTime="3060.966" EndTime="3061.787" Participant="me013">
Maybe <Emphasis> part </Emphasis> of it.
</Segment>
<Segment StartTime="3061.598" EndTime="3063.161" Participant="me011">
Make everyone read digits.
</Segment>
<Segment StartTime="3063.199" EndTime="3065.057" Participant="mn005">
<VocalSound Description="laugh"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3063.278" EndTime="3063.974" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3063.458" EndTime="3066.011" Participant="fe016">
<VocalSound Description="laugh"/> At the same time.
</Segment>
<Segment StartTime="3063.974" EndTime="3064.766" Participant="me013">
At the same time.
</Segment>
<Segment StartTime="3063.981" EndTime="3064.560" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3064.560" EndTime="3065.515" Participant="me011">
At the same time.
</Segment>
<Segment StartTime="3064.670" EndTime="3065.900" Participant="mn017">
Please. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3064.766" EndTime="3066.823" Participant="me013">
<VocalSound Description="laugh"/> Yeah.
</Segment>
<Segment StartTime="3065.057" EndTime="3066.405" Participant="mn005">
<Uncertain NumSyllables="4"> @@ </Uncertain>
</Segment>
<Segment StartTime="3066.823" EndTime="3068.148" Participant="me013">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3067.892" EndTime="3068.363" Participant="fe016">
We c-
</Segment>
<Segment StartTime="3068.148" EndTime="3069.355" Participant="me013">
I don't know. Any-
</Segment>
<Segment StartTime="3068.987" EndTime="3070.659" Participant="fe016">
That's their initiation into our
</Segment>
<Segment StartTime="3071.040" EndTime="3072.562" Participant="me011">
Into our - our - our cult.
</Segment>
<Segment StartTime="3071.075" EndTime="3072.598" Participant="me026">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3071.435" EndTime="3071.785" Participant="fe016">
w-
</Segment>
<Segment StartTime="3071.785" EndTime="3073.190" Participant="fe016">
Yeah, our - Yeah, our - <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="3072.562" EndTime="3073.037" Participant="me011">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3072.904" EndTime="3077.750" Participant="mn017">
Maybe the sections that are not right afte- you know, after lunch when everybody's still munching and -
</Segment>
<Segment StartTime="3074.337" EndTime="3075.069" Participant="fe016">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3077.922" EndTime="3080.131" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3078.267" EndTime="3080.093" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3078.620" EndTime="3080.125" Participant="mn017">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3079.874" EndTime="3080.547" Participant="mn014">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3080.131" EndTime="3081.078" Participant="me013">
O_K. Well -
</Segment>
<Segment StartTime="3080.528" EndTime="3083.361" Participant="fe016">
So can you send out a schedule once you know it, jus- ?
</Segment>
<Segment StartTime="3082.643" EndTime="3083.078" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="3083.516" EndTime="3086.173" Participant="me013">
O_K. Yeah. I guess I sent it around a little bit. But -
</Segment>
<Segment StartTime="3083.715" EndTime="3085.430" Participant="fe016">
Is - is there a r- ? There's a res-
</Segment>
<Segment StartTime="3085.804" EndTime="3087.580" Participant="fe016">
Is it <Emphasis> changed </Emphasis> now, or - ?
</Segment>
<Segment StartTime="3087.508" EndTime="3090.258" Participant="me013">
I hadn't heard back from Mari after I - I u- u-
</Segment>
<Segment StartTime="3090.258" EndTime="3093.561" Participant="me013">
uh, brought up the point abou- about Andreas's schedule. So,
</Segment>
<Segment StartTime="3093.561" EndTime="3094.115" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3094.115" EndTime="3097.841" Participant="me013">
um, maybe when I get back there'll be <Pause/> some - some mail from her.
</Segment>
<Segment StartTime="3096.046" EndTime="3096.521" Participant="fe016">
O_K.
</Segment>
<Segment StartTime="3099.338" EndTime="3102.068" Participant="fe008">
<VocalSound Description="inbreath"/> I'm looking forward to seeing your representation. That'd be, uh -
</Segment>
<Segment StartTime="3099.433" EndTime="3100.450" Participant="me013">
<VocalSound Description="mouth"/> So, I'll make a -
</Segment>
<Segment StartTime="3101.957" EndTime="3106.220" Participant="fe016">
And w- we should get <Pause/> the two meetings from y- I mean, I know about the first meeting, um,
</Segment>
<Segment StartTime="3102.365" EndTime="3103.835" Participant="fe008">
I'd like to see that. Yeah.
</Segment>
<Segment StartTime="3106.568" EndTime="3113.570" Participant="fe016">
but the other one that you did, the N_S_A one, which we <Pause/> hadn't done cuz we weren't running recognition on it, because the non-native speaker -
</Segment>
<Segment StartTime="3111.422" EndTime="3113.103" Participant="mn017">
<VocalSound Description="clears throat"/>
</Segment>
<Segment StartTime="3111.930" EndTime="3112.351" Participant="fe008">
Mm-hmm.
</Segment>
<Segment StartTime="3113.570" EndTime="3115.033" Participant="fe016">
there were <Emphasis> five </Emphasis> non-native speakers.
</Segment>
<Segment StartTime="3113.702" EndTime="3115.090" Participant="fe008">
Mm-hmm. I see. Mm-hmm.
</Segment>
<Segment StartTime="3115.033" EndTime="3115.613" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3115.613" EndTime="3117.330" Participant="fe016">
But, it would be useful for the -
</Segment>
<Segment StartTime="3118.502" EndTime="3120.513" Participant="fe016">
to <Emphasis> see </Emphasis> what we get <Pause/> with that one. So.
</Segment>
<Segment StartTime="3119.958" EndTime="3121.597" Participant="me001" CloseMic="false">
Which N_S_A meeting was that?
</Segment>
<Segment StartTime="3119.997" EndTime="3123.861" Participant="fe008">
Great. O_K. It's, uh, two thousand eleven twenty-one one thousand.
</Segment>
<Segment StartTime="3123.713" EndTime="3126.047" Participant="me001" CloseMic="false">
Oh, we did - That was the <Emphasis> last </Emphasis> one I gave you.
</Segment>
<Segment StartTime="3125.941" EndTime="3127.922" Participant="fe016">
Yeah, <Emphasis> three. </Emphasis> Right. So -
</Segment>
<Segment StartTime="3126.530" EndTime="3127.185" Participant="me001" CloseMic="false">
Mm-hmm.
</Segment>
<Segment StartTime="3127.630" EndTime="3129.431" Participant="fe008">
Great. I sent email when I finished the -
</Segment>
<Segment StartTime="3129.431" EndTime="3130.934" Participant="fe008">
that one. That was sort of son-
</Segment>
<Segment StartTime="3130.075" EndTime="3131.420" Participant="fe016">
N_S_<Emphasis> A </Emphasis> three, I think.
</Segment>
<Segment StartTime="3130.135" EndTime="3130.541" Participant="me001" CloseMic="false">
Yeah.
</Segment>
<Segment StartTime="3131.900" EndTime="3132.612" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3132.120" EndTime="3137.230" Participant="fe008">
Yeah, that's right. That's right. That's much simpler. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3132.904" EndTime="3135.860" Participant="fe016">
<VocalSound Description="laugh"/> I don't know what they said but I know the number. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="3133.730" EndTime="3136.636" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3135.579" EndTime="3144.722" Participant="me013">
Th- that part's <Emphasis> definitely </Emphasis> gonna confuse somebody who looks at these later. I mean, this is - we- we're recording secret N_S_A meetings? I mean, it's - <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3135.816" EndTime="3137.299" Participant="mn017">
<VocalSound Description="mouth"/> Right. <VocalSound Description="breath"/>
</Segment>
<Segment StartTime="3136.353" EndTime="3138.071" Participant="me001" CloseMic="false">
I remember the <Emphasis> date. </Emphasis> So.
</Segment>
<Segment StartTime="3136.614" EndTime="3139.619" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3138.226" EndTime="3142.180" Participant="mn017">
Um. <VocalSound Description="outbreath"/> Not the - <VocalSound Description="breath-laugh"/> Yeah.
</Segment>
<Segment StartTime="3140.515" EndTime="3143.797" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3140.720" EndTime="3141.729" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3140.991" EndTime="3143.203" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3141.230" EndTime="3145.300" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3143.340" EndTime="3145.635" Participant="mn017">
Uh. The - th- the -
</Segment>
<Segment StartTime="3143.797" EndTime="3145.062" Participant="fe008">
Yeah. Not <Emphasis> that </Emphasis> N_S_A.
</Segment>
<Segment StartTime="3144.722" EndTime="3149.192" Participant="me013">
It's network services and applications. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3145.062" EndTime="3149.394" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3145.100" EndTime="3151.930" Participant="fe016">
They <Emphasis> are </Emphasis> hard to understand. <VocalSound Description="laugh"/> They're very, uh, out there. I have no idea what they're talking about. <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3146.622" EndTime="3148.106" Participant="mn017">
<Uncertain> Wait. </Uncertain> The -
</Segment>
<Segment StartTime="3147.306" EndTime="3149.338" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3147.590" EndTime="3148.930" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3148.853" EndTime="3150.117" Participant="me001" CloseMic="false">
Yeah, they're cryptic.
</Segment>
<Segment StartTime="3150.805" EndTime="3152.090" Participant="mn017">
The, um -
</Segment>
<Segment StartTime="3150.859" EndTime="3152.177" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3150.952" EndTime="3151.830" Participant="me013">
Yeah. <VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3152.401" EndTime="3152.737" Participant="me013">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3152.800" EndTime="3160.134" Participant="mn017">
th- the <Emphasis> other </Emphasis> good thing about the alignments is that, um, it's not always the <Emphasis> machine's </Emphasis> fault if it doesn't work. <Comment Description="rising intonation"/> So, you can actually find, um,
</Segment>
<Segment StartTime="3159.893" EndTime="3162.878" Participant="fe016">
It's the person's fault. <VocalSound Description="laugh"/> It's Morgan's fault. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="3160.430" EndTime="3163.712" Participant="mn017">
problem - uh, proble- <VocalSound Description="laugh"/> You can find -
</Segment>
<Segment StartTime="3161.051" EndTime="3164.544" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3161.357" EndTime="3162.676" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3161.482" EndTime="3164.749" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3162.878" EndTime="3164.520" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3163.470" EndTime="3164.600" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3163.712" EndTime="3164.507" Participant="mn017">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3164.464" EndTime="3164.800" Participant="me011">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3164.507" EndTime="3169.175" Participant="mn017">
You can find, uh, problems with - with the transcripts, um, you know,
</Segment>
<Segment StartTime="3164.749" EndTime="3167.103" Participant="me013">
It's always Morgan's fault. <VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3164.895" EndTime="3165.920" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3168.558" EndTime="3168.973" Participant="me011">
Oh.
</Segment>
<Segment StartTime="3169.631" EndTime="3171.291" Participant="mn017">
and go back and fix them. But -
</Segment>
<Segment StartTime="3169.847" EndTime="3170.449" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="3170.975" EndTime="3175.405" Participant="fe016">
Tha- There are some cases like where the - the wrong speaker - uh, these ca-
</Segment>
<Segment StartTime="3175.405" EndTime="3182.073" Participant="fe016">
<Emphasis> Not </Emphasis> a <Emphasis> lot, </Emphasis> but where the - the wrong person - the - the speech is addre- attached to the wrong speaker and you can tell that when you run it.
</Segment>
<Segment StartTime="3182.073" EndTime="3182.684" Participant="fe016">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3182.684" EndTime="3189.570" Participant="fe016">
Or at least you can get <Pause/> clues to it. So these are from the early transcriptions that people did on the mixed signals, like what you have.
</Segment>
<Segment StartTime="3182.786" EndTime="3183.300" Participant="fe008">
Interesting.
</Segment>
<Segment StartTime="3184.092" EndTime="3185.654" Participant="fe008">
<VocalSound Description="inbreath"/> I guess it <Emphasis> does </Emphasis> w-
</Segment>
<Segment StartTime="3186.840" EndTime="3187.255" Participant="fe008">
Mm-hmm. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="3187.255" EndTime="3188.325" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3188.325" EndTime="3193.826" Participant="fe008">
It <Emphasis> also </Emphasis> raises the possibility of, um, using <Emphasis> that </Emphasis> kind of representation - I mean, I don't know, this'd be something we'd wanna check, <Comment Description="embedded"/>
</Segment>
<Segment StartTime="3193.826" EndTime="3199.585" Participant="fe008">
but maybe using that representation for data <Emphasis> entry </Emphasis> and then <Emphasis> displaying </Emphasis> it on the channelized, uh, representation, cuz it - I think that the -
</Segment>
<Segment StartTime="3199.585" EndTime="3205.725" Participant="fe008">
I mean, my - my <Emphasis> preference </Emphasis> in terms of, like, <Emphasis> looking </Emphasis> at the data is to see it <Pause/> in this kind of musical score format.
</Segment>
<Segment StartTime="3206.139" EndTime="3206.556" Participant="fe016">
Mm-hmm.
</Segment>
<Segment StartTime="3206.180" EndTime="3208.985" Participant="fe008">
And <Emphasis> also, </Emphasis> s- you know, <Emphasis> Sue's </Emphasis> preference as <Emphasis> well. </Emphasis> <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3208.730" EndTime="3210.490" Participant="fe016">
Yeah, if you can get it to -
</Segment>
<Segment StartTime="3208.985" EndTime="3213.178" Participant="fe008">
And - and - <Emphasis> but, </Emphasis> I mean, this - if this is a better interface for making these kinds of,
</Segment>
<Segment StartTime="3213.178" EndTime="3220.046" Participant="fe008">
uh, you know, lo- clos- local changes, then that'd be fine, too. I don't - I have no idea. I think this is something that would need to be <Emphasis> checked. </Emphasis>
</Segment>
<Segment StartTime="3221.273" EndTime="3222.499" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="3221.894" EndTime="3222.242" Participant="fe008">
Yeah.
</Segment>
<Segment StartTime="3222.499" EndTime="3229.752" Participant="me013">
<Pronounce Pronunciation="&quot;kay&quot;"> O_K. </Pronounce> Th- the <Emphasis> other </Emphasis> thing <Emphasis> I </Emphasis> had actually was, I - I didn't <Emphasis> realize </Emphasis> this till <Emphasis> today, </Emphasis> <VocalSound Description="laugh"/> but, uh, this is, uh, Jose's last <Emphasis> day. </Emphasis>
</Segment>
<Segment StartTime="3227.136" EndTime="3227.944" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3229.778" EndTime="3234.137" Participant="mn005">
Is my last - my last day. My - <VocalSound Description="laugh"/> my last <Emphasis> meeting </Emphasis> <Pause/> about meetings.
</Segment>
<Segment StartTime="3229.876" EndTime="3230.607" Participant="me011">
Yeah.
</Segment>
<Segment StartTime="3230.334" EndTime="3232.648" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3230.511" EndTime="3231.539" Participant="fe016">
Oh!
</Segment>
<Segment StartTime="3230.750" EndTime="3231.873" Participant="me001" CloseMic="false">
Oh!
</Segment>
<Segment StartTime="3230.810" EndTime="3231.740" Participant="mn017">
Oh!
</Segment>
<Segment StartTime="3231.108" EndTime="3231.690" Participant="fe008">
<Emphasis> Oh. </Emphasis>
</Segment>
<Segment StartTime="3231.596" EndTime="3232.851" Participant="me011">
You're not gonna be here tomorrow?
</Segment>
<Segment StartTime="3233.781" EndTime="3235.192" Participant="me011">
Oh, that's right. Tomorrow - <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="3234.137" EndTime="3239.294" Participant="mn005">
Yeah. Because, eh, I leave, eh, the next Sunday. I will come back to home - to Spain.
</Segment>
<Segment StartTime="3235.035" EndTime="3236.361" Participant="mn014">
The last meeting meeting?
</Segment>
<Segment StartTime="3236.410" EndTime="3237.210" Participant="me011">
It's off.
</Segment>
<Segment StartTime="3237.628" EndTime="3238.413" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="3237.987" EndTime="3238.561" Participant="fe016">
Oh.
</Segment>
<Segment StartTime="3239.294" EndTime="3240.056" Participant="mn005">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3239.399" EndTime="3241.375" Participant="me013">
Yeah. I d- so I - I jus-
</Segment>
<Segment StartTime="3239.944" EndTime="3240.300" Participant="mn017">
Mm-hmm. <Comment Description="very quietly"/>
</Segment>
<Segment StartTime="3240.056" EndTime="3250.779" Participant="mn005">
And I - I would like to - to - to say thank you very much, eh, to all people <Pause/> in the group and at ICSI, because I - I enjoyed <Uncertain NumSyllables="3"> @@ </Uncertain> very much, uh. <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3240.262" EndTime="3240.657" Participant="fe016">
Oh.
</Segment>
<Segment StartTime="3243.999" EndTime="3244.395" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="3245.499" EndTime="3246.725" Participant="me011">
Yeah. It was good having you.
</Segment>
<Segment StartTime="3245.723" EndTime="3246.561" Participant="mn017">
Mmm.
</Segment>
<Segment StartTime="3246.767" EndTime="3247.280" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="3248.674" EndTime="3249.881" Participant="mn017">
<VocalSound Description="outbreath"/> Mmm.
</Segment>
<Segment StartTime="3250.779" EndTime="3258.161" Participant="mn005">
And I'm <Emphasis> sorry </Emphasis> by the result of overlapping, because, eh, <VocalSound Description="laugh"/> I haven't good results, eh, yet but, eh,
</Segment>
<Segment StartTime="3253.421" EndTime="3256.314" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3253.611" EndTime="3253.870" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3254.054" EndTime="3254.986" Participant="mn017">
<VocalSound Description="laugh"/> <NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="3258.161" EndTime="3258.820" Participant="mn005">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="3258.820" EndTime="3263.647" Participant="mn005">
I - <VocalSound Description="mouth"/> I pretend <Comment Description="meaning &quot;intend&quot;"/> to - to continuing out to Spain, eh,
</Segment>
<Segment StartTime="3263.896" EndTime="3268.881" Participant="mn005">
during the - the following months, eh, because I have, eh, another ideas
</Segment>
<Segment StartTime="3265.729" EndTime="3266.224" Participant="me013">
Uh-huh.
</Segment>
<Segment StartTime="3268.881" EndTime="3273.806" Participant="mn005">
but, eh, I haven't enough time to - to - <VocalSound Description="laugh"/> with six months it's not enough to -
</Segment>
<Segment StartTime="3273.419" EndTime="3274.160" Participant="me011">
Yep.
</Segment>
<Segment StartTime="3273.806" EndTime="3274.753" Participant="mn005">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="3273.939" EndTime="3274.473" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="3274.753" EndTime="3276.809" Participant="mn005">
to - to research, eh, and
</Segment>
<Segment StartTime="3276.809" EndTime="3280.547" Participant="mn005">
e- i- I mean, if, eh, the topic is, eh, so difficult, uh,
</Segment>
<Segment StartTime="3278.720" EndTime="3280.639" Participant="mn017">
<VocalSound Description="clears throat"/>
</Segment>
<Segment StartTime="3280.903" EndTime="3283.137" Participant="mn005">
in my opinion, there isn't -
</Segment>
<Segment StartTime="3280.981" EndTime="3281.634" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3282.109" EndTime="3288.717" Participant="me013">
Yeah. Maybe somebody else will come along and will be, uh, interested in working on it and could start off from where you are also, you know.
</Segment>
<Segment StartTime="3283.868" EndTime="3285.490" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="3288.717" EndTime="3291.111" Participant="me013">
They'd make use of - of what you've <Emphasis> done. </Emphasis>
</Segment>
<Segment StartTime="3288.764" EndTime="3289.719" Participant="mn005">
Yeah.
</Segment>
<Segment StartTime="3291.626" EndTime="3292.477" Participant="mn005">
Yeah.
</Segment>
<Segment StartTime="3292.120" EndTime="3292.594" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="3292.793" EndTime="3297.831" Participant="mn005">
<VocalSound Description="inbreath"/> But, eh, I - I will try to recommend, eh, at, eh,
</Segment>
<Segment StartTime="3294.601" EndTime="3296.361" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="3297.831" EndTime="3298.222" Participant="mn005">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3298.222" EndTime="3299.833" Participant="mn005">
the Spanish government but, eh,
</Segment>
<Segment StartTime="3299.833" EndTime="3302.687" Participant="mn005">
the following <Uncertain NumSyllables="1"> @@ </Uncertain> scholarship, eh, eh,
</Segment>
<Segment StartTime="3302.687" EndTime="3303.150" Participant="mn005">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="3303.150" EndTime="3307.486" Participant="mn005">
eh, will be here <Pause/> <Emphasis> more </Emphasis> <Emphasis> time, </Emphasis> because eh, i- in my opinion is - is better,
</Segment>
<Segment StartTime="3307.486" EndTime="3307.998" Participant="mn005">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3307.998" EndTime="3311.237" Participant="mn005">
eh, for <Emphasis> us </Emphasis> <Pause/> to - to spend more time here
</Segment>
<Segment StartTime="3311.237" EndTime="3311.627" Participant="mn005">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3311.627" EndTime="3315.427" Participant="mn005">
and to work more time i- i- in a topic. <Uncertain> No </Uncertain>?
</Segment>
<Segment StartTime="3315.427" EndTime="3315.794" Participant="mn005">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3315.543" EndTime="3317.758" Participant="me013">
Yeah, it's a very short time. Yeah. <Uncertain> Yeah. </Uncertain> <Comment Description="last syllable, whispered"/>
</Segment>
<Segment StartTime="3315.794" EndTime="3316.389" Participant="mn005">
But, uh -
</Segment>
<Segment StartTime="3316.542" EndTime="3318.760" Participant="me011">
Yeah, six months is <Emphasis> hard. </Emphasis> I think a year
</Segment>
<Segment StartTime="3316.794" EndTime="3317.795" Participant="mn005">
Yeah. <Uncertain> It is </Uncertain>.
</Segment>
<Segment StartTime="3318.051" EndTime="3318.627" Participant="mn005">
Yeah.
</Segment>
<Segment StartTime="3318.629" EndTime="3319.650" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="3319.108" EndTime="3319.782" Participant="mn005">
It's difficult.
</Segment>
<Segment StartTime="3319.488" EndTime="3320.690" Participant="me011">
is a lot better.
</Segment>
<Segment StartTime="3319.782" EndTime="3320.212" Participant="mn005">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3320.212" EndTime="3321.644" Participant="mn005">
You - e- you have, eh -
</Segment>
<Segment StartTime="3321.644" EndTime="3328.470" Participant="mn005">
you are <Emphasis> lucky, </Emphasis> and you - you find a solution <Comment Description="while laughing"/> in - in - in some few tim- uh, months, eh?
</Segment>
<Segment StartTime="3328.470" EndTime="3331.489" Participant="mn005">
O_K. But, eh, I think it's not, eh, common.
</Segment>
<Segment StartTime="3331.489" EndTime="3331.830" Participant="mn005">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3331.830" EndTime="3337.276" Participant="mn005">
But, eh, anyway, thank you. Thank you very much. Eh, I - I bring
</Segment>
<Segment StartTime="3337.766" EndTime="3339.472" Participant="mn005">
the chocolate, eh, to -
</Segment>
<Segment StartTime="3338.917" EndTime="3342.242" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3339.021" EndTime="3344.403" Participant="me026">
<VocalSound Description="laugh"/> <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="3339.472" EndTime="3339.839" Participant="mn005">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3339.527" EndTime="3340.966" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3339.654" EndTime="3340.643" Participant="mn017">
Mmm.
</Segment>
<Segment StartTime="3339.839" EndTime="3344.224" Participant="mn005">
<VocalSound Description="laugh"/> to tear, uh, with - with you, uh.
</Segment>
<Segment StartTime="3340.027" EndTime="3340.930" Participant="fe016">
Oh.
</Segment>
<Segment StartTime="3340.062" EndTime="3341.664" Participant="fe008">
Ah. Nice.
</Segment>
<Segment StartTime="3344.697" EndTime="3349.970" Participant="mn005">
I - I hope if you need, eh, something, eh, from us in the future,
</Segment>
<Segment StartTime="3349.970" EndTime="3351.689" Participant="mn005">
I - I will be at Spain,
</Segment>
<Segment StartTime="3351.689" EndTime="3352.070" Participant="mn005">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3352.070" EndTime="3353.790" Participant="mn005">
to you help, uh.
</Segment>
<Segment StartTime="3353.564" EndTime="3354.060" Participant="mn017">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="3354.030" EndTime="3354.660" Participant="me011">
Great.
</Segment>
<Segment StartTime="3354.169" EndTime="3354.564" Participant="me013">
Well.
</Segment>
<Segment StartTime="3354.670" EndTime="3355.936" Participant="fe008">
Great. Thank you.
</Segment>
<Segment StartTime="3355.090" EndTime="3355.866" Participant="me013">
Thank you, Jose.
</Segment>
<Segment StartTime="3355.232" EndTime="3355.694" Participant="fe016">
Right.
</Segment>
<Segment StartTime="3355.246" EndTime="3357.298" Participant="mn005">
And, thank you very much.
</Segment>
<Segment StartTime="3355.481" EndTime="3356.091" Participant="me001" CloseMic="false">
Thank you.
</Segment>
<Segment StartTime="3357.364" EndTime="3358.226" Participant="mn017">
Have a good trip.
</Segment>
<Segment StartTime="3358.258" EndTime="3358.712" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3358.319" EndTime="3361.080" Participant="me013">
Yeah. <VocalSound Description="outbreath"/> Yeah.
</Segment>
<Segment StartTime="3358.679" EndTime="3359.442" Participant="mn005">
Thank you.
</Segment>
<Segment StartTime="3358.863" EndTime="3359.163" Participant="fe008">
Yeah. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="3359.050" EndTime="3359.682" Participant="mn017">
Keep in touch.
</Segment>
<Segment StartTime="3361.473" EndTime="3362.067" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="3362.067" EndTime="3363.965" Participant="me013">
O_K. I guess, uh,
</Segment>
<Segment StartTime="3364.497" EndTime="3367.776" Participant="me013">
unless somebody has something else, we'll read - read our digits and we'll get our - <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3364.651" EndTime="3365.363" Participant="me011">
Digits?
</Segment>
<Segment StartTime="3365.387" EndTime="3366.233" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="3366.547" EndTime="3368.707" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3366.673" EndTime="3370.058" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="3366.922" EndTime="3368.208" Participant="mn014">
Uh. Oops. <Comment Description="interjection"/>
</Segment>
<Segment StartTime="3367.084" EndTime="3369.715" Participant="me011">
Are we gonna do <Pronounce Pronunciation="&quot;em&quot;"> them </Pronounce> simultaneously or - ? <VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3367.776" EndTime="3371.416" Participant="me013">
get our last bit of, uh, Jose's - <Pronounce> Jose </Pronounce> - Jose's digit - Uh, I'm sorry?
</Segment>
<Segment StartTime="3368.333" EndTime="3369.522" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3369.268" EndTime="3370.043" Participant="mn005">
You - eh -
</Segment>
<Segment StartTime="3371.362" EndTime="3377.223" Participant="mn005">
Ye- ye- you prefer, eh, to eat, eh, chocolate, eh, at the coffee break, eh, at the - ?
</Segment>
<Segment StartTime="3371.638" EndTime="3372.054" Participant="fe016">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3377.223" EndTime="3377.610" Participant="mn005">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3377.610" EndTime="3380.042" Participant="mn005">
Or you prefer <Emphasis> now, </Emphasis> before after - ?
</Segment>
<Segment StartTime="3378.030" EndTime="3380.594" Participant="fe008">
<VocalSound Description="inbreath"/> Well, we have a time - <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="3378.621" EndTime="3380.819" Participant="mn017">
No, we prefer to keep it for ourselves.
</Segment>
<Segment StartTime="3380.042" EndTime="3380.601" Participant="mn005">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3380.492" EndTime="3383.470" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3380.594" EndTime="3384.155" Participant="fe008">
<VocalSound Description="laugh"/> Well, we have a s- a time - <Emphasis> time </Emphasis> constraint. Yeah.
</Segment>
<Segment StartTime="3380.601" EndTime="3385.922" Participant="mn005">
<VocalSound Description="laugh"/> Yeah. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="3380.760" EndTime="3381.692" Participant="mn014">
During -
</Segment>
<Segment StartTime="3380.793" EndTime="3386.892" Participant="me013">
<VocalSound Description="laugh"/> So keep it away from that end of the table. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3380.819" EndTime="3385.269" Participant="mn017">
Yeah, yeah. <VocalSound Description="laugh"/> Yeah.
</Segment>
<Segment StartTime="3380.836" EndTime="3386.099" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3381.179" EndTime="3383.315" Participant="me026">
<VocalSound Description="laugh"/> <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="3381.692" EndTime="3386.270" Participant="mn014">
<VocalSound Description="laugh"/> during digits. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3384.448" EndTime="3384.824" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3385.269" EndTime="3391.256" Participant="mn017">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3385.856" EndTime="3386.508" Participant="fe008">
Yeah.
</Segment>
<Segment StartTime="3385.922" EndTime="3390.160" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3386.099" EndTime="3388.430" Participant="fe016">
Why is it that I can read your mind?
</Segment>
<Segment StartTime="3386.229" EndTime="3389.374" Participant="me011">
Well, we've gotta wait until after di- after we take the mikes off.
</Segment>
<Segment StartTime="3388.100" EndTime="3391.265" Participant="mn014">
No, no. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3388.316" EndTime="3389.823" Participant="me026">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3389.521" EndTime="3391.638" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3390.710" EndTime="3392.911" Participant="me011">
So are we gonna do digits simultaneously or what?
</Segment>
<Segment StartTime="3391.316" EndTime="3393.019" Participant="me013">
Well? <Pause/> Yeah.
</Segment>
<Segment StartTime="3391.638" EndTime="3394.811" Participant="fe016">
You - This is our reward if we <Pause/> do our digi-
</Segment>
<Segment StartTime="3392.195" EndTime="3392.808" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="3392.229" EndTime="3392.821" Participant="fe008">
O_K.
</Segment>
<Segment StartTime="3394.080" EndTime="3400.352" Participant="mn005">
I - I think, eh, it's enough, eh, for more peopl- for more people <Pause/> after. But, eh -
</Segment>
<Segment StartTime="3394.122" EndTime="3396.737" Participant="mn014">
Simultaneous digit chocolate task.
</Segment>
<Segment StartTime="3395.380" EndTime="3397.971" Participant="me013">
We're gonna - we're gonna do digits at the same -
</Segment>
<Segment StartTime="3395.633" EndTime="3396.653" Participant="fe016">
<VocalSound Description="mouth"/> Oh.
</Segment>
<Segment StartTime="3396.651" EndTime="3398.633" Participant="me001" CloseMic="false">
I'm going to get a ticket if people don't eat some.
</Segment>
<Segment StartTime="3397.755" EndTime="3398.941" Participant="mn017">
Mmm! <Comment Description="lengthened"/> <Comment Description="vocal gesture, commenting on chocolates"/>
</Segment>
<Segment StartTime="3398.237" EndTime="3399.423" Participant="mn014">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3398.869" EndTime="3400.008" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3399.459" EndTime="3400.013" Participant="me011">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3399.836" EndTime="3401.783" Participant="fe008">
That's nice. Wow.
</Segment>
<Segment StartTime="3400.169" EndTime="3400.861" Participant="mn017">
Mm-hmm.
</Segment>
<Segment StartTime="3400.226" EndTime="3401.634" Participant="me001" CloseMic="false">
<VocalSound Description="laugh"/> So, I have a vested interest.
</Segment>
<Segment StartTime="3400.505" EndTime="3402.680" Participant="me013">
Um. <Pause/> Well -
</Segment>
<Segment StartTime="3400.799" EndTime="3402.167" Participant="fe016">
Oh, thanks, Jose.
</Segment>
<Segment StartTime="3401.469" EndTime="3407.362" Participant="mn005">
To Andreas, the idea is - is good. <VocalSound Description="laugh"/> s- To eat here. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3401.987" EndTime="3402.758" Participant="mn017">
Mmm.
</Segment>
<Segment StartTime="3403.406" EndTime="3404.915" Participant="fe008">
Wow. Very nice.
</Segment>
<Segment StartTime="3403.470" EndTime="3404.670" Participant="mn017">
Oh.
</Segment>
<Segment StartTime="3403.830" EndTime="3404.738" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3405.046" EndTime="3408.786" Participant="me013">
Tha- that's - that looks <Emphasis> great. </Emphasis> Alright, so in the interest of getting to the -
</Segment>
<Segment StartTime="3405.076" EndTime="3405.929" Participant="fe016">
Oh, wow.
</Segment>
<Segment StartTime="3405.764" EndTime="3408.751" Participant="mn017">
Oh, yeah. Th- it doesn't - it won't leave this room. <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3408.319" EndTime="3412.201" Participant="fe016">
We could do digits while other people eat. So it's background crunching.
</Segment>
<Segment StartTime="3409.543" EndTime="3411.680" Participant="mn014">
Yeah. Yeah. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3410.106" EndTime="3410.699" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="3410.872" EndTime="3412.596" Participant="mn017">
<VocalSound Description="breath"/> Mmm. <Comment Description="vocal gesture, commenting on chocolates"/>
</Segment>
<Segment StartTime="3410.925" EndTime="3412.430" Participant="mn005">
Yeah. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="3412.430" EndTime="3418.859" Participant="mn005">
Is, eh, a - another acoustic event. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3412.560" EndTime="3413.229" Participant="fe008">
Nice.
</Segment>
<Segment StartTime="3412.679" EndTime="3422.464" Participant="me026">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3412.733" EndTime="3414.288" Participant="fe016">
We don't have background chewing.
</Segment>
<Segment StartTime="3413.245" EndTime="3416.509" Participant="mn014">
Background crunch. Yeah. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3414.565" EndTime="3416.862" Participant="fe016">
No, we don't have any data with background eating.
</Segment>
<Segment StartTime="3414.784" EndTime="3415.595" Participant="mn017">
Mmm. <Comment Description="vocal gesture, commenting on chocolates"/>
</Segment>
<Segment StartTime="3416.929" EndTime="3419.268" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3417.100" EndTime="3419.790" Participant="mn014">
Yeah. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3417.144" EndTime="3418.707" Participant="mn017">
<VocalSound Description="outbreath"/> <VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3417.740" EndTime="3419.955" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3418.323" EndTime="3419.428" Participant="fe016">
I'm serious. You-
</Segment>
<Segment StartTime="3419.268" EndTime="3422.529" Participant="me013">
She's - she's serious. <VocalSound Description="breath-laugh"/> She is serious.
</Segment>
<Segment StartTime="3419.690" EndTime="3424.221" Participant="me011">
It's just the <Emphasis> rest </Emphasis> of the digits - the rest of the digits are very <Emphasis> clean, </Emphasis>
</Segment>
<Segment StartTime="3420.123" EndTime="3421.381" Participant="fe016">
I am serious. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="3420.659" EndTime="3422.340" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3422.109" EndTime="3422.702" Participant="mn017">
Mmm. <Comment Description="vocal gesture, commenting on chocolates"/>
</Segment>
<Segment StartTime="3423.708" EndTime="3425.035" Participant="mn005">
Are you - ? <Uncertain> Oh, <Emphasis> they're </Emphasis> clean. </Uncertain> <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="3424.217" EndTime="3424.794" Participant="fe016">
Well - ?
</Segment>
<Segment StartTime="3424.221" EndTime="3428.315" Participant="me011">
um, without a lot of background noise, so I'm just not sure -
</Segment>
<Segment StartTime="3424.619" EndTime="3425.014" Participant="mn014">
<Uncertain> Yeah </Uncertain>!
</Segment>
<Segment StartTime="3425.287" EndTime="3425.735" Participant="fe016">
And it -
</Segment>
<Segment StartTime="3426.646" EndTime="3436.459" Participant="fe016">
You have to write down, like, while y- what you're - what ch- chocolate you're eating cuz they might make different sounds, like n- nuts - chocolate with nuts, chocolate without nuts. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="3428.635" EndTime="3429.312" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3430.289" EndTime="3432.089" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3431.326" EndTime="3435.728" Participant="mn017">
<VocalSound Description="laugh"/> Chocolate adaptation.
</Segment>
<Segment StartTime="3431.349" EndTime="3433.624" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3431.609" EndTime="3436.851" Participant="me026">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3432.348" EndTime="3433.059" Participant="fe008">
Oh.
</Segment>
<Segment StartTime="3432.633" EndTime="3434.097" Participant="mn014">
Crunchy frogs.
</Segment>
<Segment StartTime="3432.760" EndTime="3433.549" Participant="me013">
Um -
</Segment>
<Segment StartTime="3434.620" EndTime="3440.690" Participant="me013">
Actually - <VocalSound Description="inbreath"/> actually kind of careful cuz I have a strong allergy to nuts, so I have to sort of figure out one without th-
</Segment>
<Segment StartTime="3436.034" EndTime="3436.865" Participant="mn005">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3439.785" EndTime="3441.370" Participant="fe016">
That w- Oh, <Emphasis> yeah, </Emphasis> they - they might.
</Segment>
<Segment StartTime="3441.301" EndTime="3443.163" Participant="me013">
It's hard to - hard to say.
</Segment>
<Segment StartTime="3443.121" EndTime="3444.888" Participant="fe016">
Maybe <Emphasis> those? </Emphasis> They're so -
</Segment>
<Segment StartTime="3444.369" EndTime="3447.237" Participant="me013">
I don't know. <VocalSound Description="breath"/> Um -
</Segment>
<Segment StartTime="3444.888" EndTime="3445.638" Participant="fe016">
I don't know. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="3445.610" EndTime="3447.483" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3446.372" EndTime="3447.660" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3447.613" EndTime="3447.989" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="3447.660" EndTime="3454.770" Participant="fe016">
This is - You know, this is a different kind of speech, looking at chocolates, deciding - you know, it's another <Emphasis> style. </Emphasis> <VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3447.989" EndTime="3448.622" Participant="me013">
Well -
</Segment>
<Segment StartTime="3449.108" EndTime="3450.275" Participant="mn005">
Take - take several.
</Segment>
<Segment StartTime="3450.223" EndTime="3451.152" Participant="mn017">
Mmm. <Comment Description="vocal gesture, commenting on chocolates"/>
</Segment>
<Segment StartTime="3450.275" EndTime="3453.335" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3450.947" EndTime="3451.659" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3451.941" EndTime="3452.688" Participant="mn017">
Mmm. <Comment Description="vocal gesture, commenting on chocolates"/>
</Segment>
<Segment StartTime="3452.199" EndTime="3455.840" Participant="me013">
Yeah. I may - I may hold off. But if I was - eh, but maybe I'll get some later.
</Segment>
<Segment StartTime="3455.840" EndTime="3456.572" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3455.982" EndTime="3456.953" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3456.023" EndTime="3456.518" Participant="me011">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3456.038" EndTime="3456.829" Participant="mn017">
Mmm. <Comment Description="vocal gesture, commenting on chocolates"/>
</Segment>
<Segment StartTime="3456.572" EndTime="3460.910" Participant="me013">
Thanks. Well - well, why don't we - ? He - he's worried about a <Emphasis> ticket. </Emphasis> Why don't we do a simultaneous one?
</Segment>
<Segment StartTime="3458.857" EndTime="3460.458" Participant="mn005">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="3460.343" EndTime="3461.361" Participant="me001" CloseMic="false">
Yeah. Thank you!
</Segment>
<Segment StartTime="3460.440" EndTime="3460.853" Participant="fe016">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="3460.853" EndTime="3461.402" Participant="fe016">
O_K.
</Segment>
<Segment StartTime="3460.869" EndTime="3461.621" Participant="me011">
<Pronounce Pronunciation="&quot;nn-okay&quot;"> O_K. </Pronounce> <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="3461.374" EndTime="3461.786" Participant="fe008">
<Pronounce Pronunciation="&quot;kay&quot;"> O_K. </Pronounce>
</Segment>
<Segment StartTime="3461.402" EndTime="3462.042" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3461.442" EndTime="3463.242" Participant="mn014">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="3461.455" EndTime="3461.956" Participant="mn017">
Mmm. <Comment Description="vocal gesture, commenting on chocolates"/>
</Segment>
<Segment StartTime="3461.581" EndTime="3463.007" Participant="me013">
Simultaneous one? O_K.
</Segment>
<Segment StartTime="3462.214" EndTime="3464.831" Participant="me011">
Remember to read the transcript number, please.
</Segment>
<Segment StartTime="3462.457" EndTime="3465.000" Participant="fe016">
And you laughed at me, too, f- the first time I said that.
</Segment>
<Segment StartTime="3462.589" EndTime="3463.130" Participant="mn017">
Right. <Comment Description="with mouth full"/>
</Segment>
<Segment StartTime="3462.639" EndTime="3464.325" Participant="mn005">
O_K. <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3463.242" EndTime="3464.246" Participant="mn014">
<VocalSound Description="throat clearing"/>
</Segment>
<Segment StartTime="3464.246" EndTime="3465.813" Participant="mn014">
Oops. <Comment Description="interjection"/> <VocalSound Description="breath"/>
</Segment>
<Segment StartTime="3464.325" EndTime="3466.035" Participant="mn005">
Yeah. <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="3464.392" EndTime="3465.230" Participant="me013">
I have to <Emphasis> what? </Emphasis>
</Segment>
<Segment StartTime="3465.000" EndTime="3467.452" Participant="fe016">
<VocalSound Description="inbreath"/> You laughed at me, too, the first time I sa- said -
</Segment>
<Segment StartTime="3466.740" EndTime="3470.164" Participant="me013">
<VocalSound Description="inbreath"/> I did, and now I love it so much. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3467.452" EndTime="3469.246" Participant="fe016">
You really shouldn't, uh, te-
</Segment>
<Segment StartTime="3469.461" EndTime="3470.934" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3469.530" EndTime="3470.750" Participant="me011">
O_K, everyone ready? <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="3469.589" EndTime="3470.142" Participant="fe016">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3470.702" EndTime="3477.724" Participant="fe016">
You have to sort of, um - Jose, if you haven't done this, you have to plug your ears while you're t- talking so that you don't get confused, I guess.
</Segment>
<Segment StartTime="3470.963" EndTime="3473.139" Participant="mn005">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="3471.038" EndTime="3473.340" Participant="me011" DigitTask="true">
Transcript L_ four four. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3473.112" EndTime="3473.487" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="3473.487" EndTime="3474.666" Participant="me013">
W- wait - wait a minute - wait a minute.
</Segment>
<Segment StartTime="3473.952" EndTime="3477.453" Participant="mn005">
<VocalSound Description="breath"/> <NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="3475.125" EndTime="3475.918" Participant="me013">
W- we want - we want -
</Segment>
<Segment StartTime="3476.707" EndTime="3478.043" Participant="me013">
we want it <Emphasis> synchronized. </Emphasis>
</Segment>
<Segment StartTime="3477.453" EndTime="3478.125" Participant="mn005">
Yeah.
</Segment>
<Segment StartTime="3477.659" EndTime="3479.289" Participant="fe008">
Hey, <Emphasis> you've </Emphasis> done this before. Haven't you? <Comment Description="strong rise-fall-rise contour"/>
</Segment>
<Segment StartTime="3477.724" EndTime="3480.635" Participant="fe016">
Yeah. Oh, you've done this one before? Together?
</Segment>
<Segment StartTime="3478.780" EndTime="3479.632" Participant="mn014">
That's -
</Segment>
<Segment StartTime="3479.712" EndTime="3482.998" Participant="fe008">
You've read <Pause/> digits together with us, haven't you - I mean, at the same time?
</Segment>
<Segment StartTime="3483.160" EndTime="3483.771" Participant="mn005">
No.
</Segment>
<Segment StartTime="3483.629" EndTime="3485.085" Participant="fe008">
Oh, you haven't! Oh, O_K.
</Segment>
<Segment StartTime="3483.652" EndTime="3488.259" Participant="fe016">
I'm not - we - we - Oh, and you haven't done this either. I- the first time is <Pause/> traumatic, but -
</Segment>
<Segment StartTime="3483.771" EndTime="3487.847" Participant="mn005">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="3483.989" EndTime="3484.721" Participant="me013">
O_K.
</Segment>
<Segment StartTime="3484.188" EndTime="3485.135" Participant="mn014">
Oh, yeah.
</Segment>
<Segment StartTime="3485.581" EndTime="3486.175" Participant="me013">
<VocalSound Description="inbreath"/> We-
</Segment>
<Segment StartTime="3487.465" EndTime="3489.220" Participant="me013">
Y- <VocalSound Description="outbreath"/> Yeah, bu-
</Segment>
<Segment StartTime="3488.240" EndTime="3492.332" Participant="fe008">
<VocalSound Description="inbreath"/> Oh, and the <Emphasis> groupings </Emphasis> are important, so yo- you're supposed to pause between the groupings.
</Segment>
<Segment StartTime="3488.242" EndTime="3492.318" Participant="me026">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3488.540" EndTime="3489.232" Participant="mn005">
Mmm.
</Segment>
<Segment StartTime="3488.597" EndTime="3488.934" Participant="fe016">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="3490.360" EndTime="3492.372" Participant="mn005">
The grouping. <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3492.059" EndTime="3494.600" Participant="me013">
Yeah. O_K. So, uh -
</Segment>
<Segment StartTime="3492.372" EndTime="3494.019" Participant="mn005">
Yeah. <VocalSound Description="outbreath"/>
</Segment>
<Segment StartTime="3493.374" EndTime="3495.784" Participant="mn017">
You mean that the - the <Emphasis> grouping </Emphasis> is supposed to be <Emphasis> synchronized? </Emphasis>
</Segment>
<Segment StartTime="3494.630" EndTime="3495.125" Participant="fe016">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="3495.575" EndTime="3498.091" Participant="mn005">
<VocalSound Description="laugh"/> <Comment Description="some words blurred in at end"/>
</Segment>
<Segment StartTime="3495.824" EndTime="3497.723" Participant="me011">
Yeah, sure. <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3495.909" EndTime="3499.460" Participant="fe008">
No. <VocalSound Description="laugh"/> No. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="3496.084" EndTime="3499.368" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3496.133" EndTime="3498.607" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3496.141" EndTime="3499.592" Participant="me013">
No, no. <VocalSound Description="laugh"/> Synchronized digits.
</Segment>
<Segment StartTime="3496.330" EndTime="3497.141" Participant="mn017">
No?
</Segment>
<Segment StartTime="3497.697" EndTime="3498.788" Participant="me001" CloseMic="false">
Oh, <Emphasis> wow. </Emphasis>
</Segment>
<Segment StartTime="3497.863" EndTime="3498.555" Participant="mn017">
No? <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="3498.607" EndTime="3502.705" Participant="fe016">
That'd be good. We- we'll give everybody the same sheet but they say different - <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3499.460" EndTime="3503.769" Participant="mn017">
It's like a - like a Greek - like a Greek choir? You know? Like - <Uncertain> Yeah. </Uncertain>
</Segment>
<Segment StartTime="3500.699" EndTime="3501.233" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3502.314" EndTime="3503.535" Participant="me001" CloseMic="false">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3502.609" EndTime="3507.230" Participant="me011">
<VocalSound Description="mouth"/> Hey, what a good idea. We could do the same sheet for everyone. Have <Pronounce Pronunciation="&quot;em&quot;"> them </Pronounce> all read <Pronounce Pronunciation="&quot;em&quot;"> them </Pronounce> at once.
</Segment>
<Segment StartTime="3502.810" EndTime="3503.621" Participant="me013">
Yes.
</Segment>
<Segment StartTime="3504.165" EndTime="3505.609" Participant="mn017">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3506.045" EndTime="3508.367" Participant="fe016">
Well, different digits but same groupings.
</Segment>
<Segment StartTime="3506.497" EndTime="3507.071" Participant="mn014">
Eh -
</Segment>
<Segment StartTime="3508.490" EndTime="3509.260" Participant="mn014">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="3508.751" EndTime="3511.441" Participant="me011">
Or - or just same digits. See if anyone notices. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="3509.269" EndTime="3511.080" Participant="fe016">
So they would all be - Yeah.
</Segment>
<Segment StartTime="3510.026" EndTime="3510.579" Participant="me013">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="3510.097" EndTime="3512.764" Participant="fe008">
Yeah. That'd be good. And then - then we can <Emphasis> sing </Emphasis> them next time. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3510.579" EndTime="3513.878" Participant="me013">
There's <Emphasis> so </Emphasis> many possibilities. <VocalSound Description="laugh"/> Uh. O_K, why don't we go?
</Segment>
<Segment StartTime="3512.032" EndTime="3512.624" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3512.588" EndTime="3513.733" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3513.397" EndTime="3514.448" Participant="fe008">
<VocalSound Description="breath-laugh"/> O_K.
</Segment>
<Segment StartTime="3513.733" EndTime="3514.893" Participant="fe016">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="3513.878" EndTime="3517.221" Participant="me013">
Uh, one two three - Go!
</Segment>
<Segment StartTime="3516.822" EndTime="3517.969" Participant="me011" DigitTask="true">
L_ forty four, <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3517.020" EndTime="3519.771" Participant="fe008" DigitTask="true">
Transcript L_ twenty-nine. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3517.380" EndTime="3519.407" Participant="me026" DigitTask="true">
Transcript L_ dash four one. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3517.568" EndTime="3519.404" Participant="me013" DigitTask="true">
Transcript L_ forty-six. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3517.803" EndTime="3521.950" Participant="fe016" DigitTask="true">
Four, eight eight seven, two six, one one two, one. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3517.870" EndTime="3519.812" Participant="mn014" DigitTask="true">
Transcript L_ forty three. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3517.950" EndTime="3520.360" Participant="mn017" DigitTask="true">
Transcript L_ forty five. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3518.347" EndTime="3522.350" Participant="me011" DigitTask="true">
Six eight six, nine three two, nine two <Emphasis> three. </Emphasis> <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3519.407" EndTime="3523.887" Participant="me026" DigitTask="true">
Two, eight six eight, two three, two five three, <VocalSound Description="mouth"/> eight. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3519.485" EndTime="3523.937" Participant="mn005" DigitTask="true">
s- seven nine six seven, eight, four eight five, <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3519.812" EndTime="3524.546" Participant="mn014" DigitTask="true">
Four, zero three six, three five, one four three, one. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3519.971" EndTime="3524.105" Participant="me013" DigitTask="true">
Nine six one, eight five, nine seven seven five. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3520.402" EndTime="3524.253" Participant="fe008" DigitTask="true">
Zero five five three, two, one nine zero. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3520.800" EndTime="3526.270" Participant="mn017" DigitTask="true">
Forty nine, seventy seven, one six, zero three, five zero. <Comment Description="corrected below"/> <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3522.207" EndTime="3526.429" Participant="fe016" DigitTask="true">
Nine three two six, six five, nine three, three four one one. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3523.130" EndTime="3528.210" Participant="me011" DigitTask="true">
Seven one O_ eight, one nine eight four, five eight seven <Emphasis> seven. </Emphasis> <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3523.937" EndTime="3529.394" Participant="mn005" DigitTask="true">
seven four one nine, two six eight nine, four nine seven one. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3524.297" EndTime="3528.069" Participant="me026" DigitTask="true">
Zero four eight, nine four three, two three one. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3524.570" EndTime="3528.267" Participant="me013" DigitTask="true">
Two nine nine, nine four five, three four nine three. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3524.642" EndTime="3527.955" Participant="fe008" DigitTask="true">
Eight two seven, two one six, four three six. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3525.254" EndTime="3531.060" Participant="mn014" DigitTask="true">
Three nine <Emphasis> five </Emphasis> nine, five nine one eight, five one seven five. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3526.749" EndTime="3528.140" Participant="mn017" DigitTask="true">
I'm starting over. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3526.800" EndTime="3530.798" Participant="fe016" DigitTask="true">
Seven, five six two, six nine, three three four, four. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3528.420" EndTime="3535.050" Participant="mn017" DigitTask="true">
Four nine, seven seven, one six, zero three, five zero, <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3528.490" EndTime="3531.970" Participant="me011" DigitTask="true">
Two eight four three, three six <Emphasis> two </Emphasis> three, O_ eight five <Emphasis> four. </Emphasis> <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3528.618" EndTime="3533.163" Participant="me026" DigitTask="true">
Zero six six, five four, two four three zero. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3528.670" EndTime="3532.880" Participant="fe008" DigitTask="true">
Five nine one zero, zero nine six six, eight zero eight eight. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3528.830" EndTime="3531.895" Participant="me013" DigitTask="true">
Nine three nine, six zero five, four seven seven. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3529.394" EndTime="3534.352" Participant="mn005" DigitTask="true">
three, two five four, six four, one, five six eight, <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3531.179" EndTime="3534.731" Participant="fe016" DigitTask="true">
Four one two six, four one five four, three four seven three. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3531.435" EndTime="3535.030" Participant="mn014" DigitTask="true">
Six eight seven, eight five eight, eight one four. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3532.332" EndTime="3535.504" Participant="me011" DigitTask="true">
Three one three, O_ nine six, nine four <Emphasis> nine. </Emphasis> <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3532.512" EndTime="3536.544" Participant="me013" DigitTask="true">
Five three, seven three, six seven, seven five zero seven. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3533.040" EndTime="3536.423" Participant="fe008" DigitTask="true">
Five six one, four nine six, three six four six. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3533.710" EndTime="3538.524" Participant="me026" DigitTask="true">
Seven, seven nine two, six nine, four five zero, four. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3534.861" EndTime="3539.900" Participant="mn005" DigitTask="true">
two zero nine seven, seven nine nine seven, seven eight seven eight. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3535.145" EndTime="3540.100" Participant="mn014" DigitTask="true">
Five, two seven one, six one, zero five <Emphasis> eight, </Emphasis> one. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3535.274" EndTime="3539.922" Participant="fe016" DigitTask="true">
Three six one, three six nine, Oops. <Comment Description="interjection"/> <Comment Description="corrected"/> Three <Emphasis> seven </Emphasis> nine, four O_ seven. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3535.455" EndTime="3541.616" Participant="mn017" DigitTask="true">
five O_ three, three two five, five two zero nine, <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3535.990" EndTime="3539.830" Participant="me011" DigitTask="true">
Six four seven seven, two zero, eight two, four <Emphasis> seven. </Emphasis> <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3536.711" EndTime="3539.365" Participant="fe008" DigitTask="true">
Zero six eight two, eight, six one nine. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3537.590" EndTime="3540.800" Participant="me013" DigitTask="true">
Two two eight eight, one eight four seven, one one two seven. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3539.210" EndTime="3543.188" Participant="me026" DigitTask="true">
Nine zero nine, <Emphasis> three </Emphasis> nine, <Emphasis> two </Emphasis> five two five. <Comment Description="change in voice quality"/> <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3539.663" EndTime="3543.261" Participant="fe008" DigitTask="true">
Seven eight three zero, eight five eight one, nine seven two eight. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3540.170" EndTime="3542.820" Participant="me011" DigitTask="true">
One five five, two seven, three eight nine <Emphasis> two. </Emphasis> <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3540.290" EndTime="3544.272" Participant="fe016" DigitTask="true">
Eight one, nine six, nine O_, three six, three six. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3540.424" EndTime="3546.973" Participant="mn005" DigitTask="true">
Three zero three three, zero four four five, one six zero fi- six. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3540.800" EndTime="3544.779" Participant="mn014" DigitTask="true">
Seven four one, zero five five, nine eight two. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3541.440" EndTime="3545.039" Participant="me013" DigitTask="true">
Seven one one eight, four one seven five, five five zero one. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3542.546" EndTime="3549.580" Participant="mn017" DigitTask="true">
two seven, two four, two nine, six three, five six, <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3543.188" EndTime="3543.888" Participant="me026">
<VocalSound Description="cough"/>
</Segment>
<Segment StartTime="3543.440" EndTime="3547.292" Participant="me011" DigitTask="true">
Nine O_ five seven, six eight four five, one five three eight, <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3543.637" EndTime="3546.342" Participant="fe008" DigitTask="true">
Eight four eight nine, one, one five six. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3544.272" EndTime="3547.982" Participant="fe016" DigitTask="true">
One four two two, seven four three one, nine nine three three. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3544.370" EndTime="3549.157" Participant="me026" DigitTask="true">
Seven, three five one, three four, two three one, zero. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3545.450" EndTime="3551.050" Participant="mn014" DigitTask="true">
Six nine, one four, zero five, seven nine, five six, nine two. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3546.040" EndTime="3549.273" Participant="me013" DigitTask="true">
One seven three, four eight, zero seven nine three. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3546.342" EndTime="3547.229" Participant="fe008" DigitTask="true">
Twenty-s- <Comment Description="corrected below"/> <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3547.229" EndTime="3547.909" Participant="fe008" DigitTask="true">
Oh, sorry. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3547.486" EndTime="3551.296" Participant="mn005" DigitTask="true">
Two one eight, seven one six, five one eight, <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3548.100" EndTime="3551.960" Participant="me011" DigitTask="true">
two nine, eight four, eight two, O_ three, five two. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3548.261" EndTime="3552.170" Participant="fe008" DigitTask="true">
<Emphasis> Two </Emphasis> seven, five one, zero two, five zero, two two. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3548.622" EndTime="3552.735" Participant="fe016" DigitTask="true">
Eight four seven two, O_ seven two three, O_ one seven O_. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3549.780" EndTime="3555.196" Participant="me026" DigitTask="true">
<Emphasis> Nine </Emphasis> seven, <Emphasis> three </Emphasis> seven, five nine, seven seven, two three. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3550.124" EndTime="3553.747" Participant="me013" DigitTask="true">
Four eight four three, four seven five one, eight five nine eight. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3550.250" EndTime="3557.700" Participant="mn017" DigitTask="true">
eight two nine zero, six four three one, six eight five five, <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3551.280" EndTime="3555.543" Participant="mn014" DigitTask="true">
One nine, one six, five four, eight four, three one. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3551.562" EndTime="3556.063" Participant="mn005" DigitTask="true">
four eh five eh seven, six eight, eh nine zero one two <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3555.840" EndTime="3560.620" Participant="me026" DigitTask="true">
Two nine eight, one two five, two two nine nine. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3555.880" EndTime="3559.740" Participant="mn014" DigitTask="true">
Seven seven five, zero eight three, nine five four seven. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3556.063" EndTime="3561.392" Participant="mn005" DigitTask="true">
t- two two five eight, five two eight nine, five seven seven eight. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3558.560" EndTime="3564.510" Participant="mn017" DigitTask="true">
four six nine, one five six, four seven eight one, <Comment Description="while laughing"/> <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3562.653" EndTime="3563.687" Participant="fe008">
Mmm! <Comment Description="vocal gesture, commenting on chocolates"/>
</Segment>
<Segment StartTime="3563.093" EndTime="3563.824" Participant="fe016">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="3563.379" EndTime="3564.510" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3564.550" EndTime="3568.467" Participant="mn017" DigitTask="true">
five four eight seven, three, five five five, <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3565.292" EndTime="3565.782" Participant="mn005">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="3568.686" EndTime="3572.525" Participant="mn017" DigitTask="true">
one six three one, two, eight three eight, <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3571.046" EndTime="3572.648" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3573.110" EndTime="3578.020" Participant="mn017" DigitTask="true">
eight zero nine, zero seven four, one six four seven. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="3577.794" EndTime="3578.111" Participant="me013">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="3577.989" EndTime="3579.373" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3578.020" EndTime="3580.892" Participant="me011">
<VocalSound Description="laugh"/> Did you read it twice or what? <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3578.111" EndTime="3586.171" Participant="me013">
And Andreas has the last word. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3578.800" EndTime="3580.894" Participant="me026">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="3579.865" EndTime="3583.904" Participant="fe016">
He's try- No, he's trying to get good recognition performance. <Comment Description="not spoken into mike"/> <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3579.884" EndTime="3580.896" Participant="mn005">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3579.888" EndTime="3581.242" Participant="fe008">
<VocalSound Description="laugh"/> He had the h-
</Segment>
<Segment StartTime="3582.440" EndTime="3585.940" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3582.649" EndTime="3585.229" Participant="mn005">
Yeah. Yeah. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="3582.771" EndTime="3585.277" Participant="fe008">
<VocalSound Description="laugh"/> He had the - the long form.
</Segment>
<Segment StartTime="3583.160" EndTime="3584.260" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="3583.904" EndTime="3585.830" Participant="fe016">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="3584.715" EndTime="3585.268" Participant="mn017">
No.
</Segment>
<Segment StartTime="3584.920" EndTime="3586.171" Participant="me011">
And we're off.
</Segment>
</Transcript>
</Meeting>