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="Bmr013" DateTimeStamp="2001-02-08-1430" Version="2" VersionDate="Jan 7 2004">
<Preamble>
<Notes>
Transcription begins at 1:38 into recording. Conversation earlier is
segmented, with pieces marked simply as "pre-meeting utterance",
during the period that the recording equipment was being set up
and debugged.
</Notes>
<Channels>
<Channel Name="chan1" Mic="s1" Transmission="w1" Gain="8" AudioFile="chan1.sph"/>
<Channel Name="chan2" Mic="s1" Transmission="w1" Gain="8" AudioFile="chan2.sph"/>
<Channel Name="chan3" Mic="s1" Transmission="w1" Gain="8" AudioFile="chan3.sph"/>
<Channel Name="chan4" Mic="s1" Transmission="w1" Gain="8" AudioFile="chan4.sph"/>
<Channel Name="chan5" Mic="a1" Transmission="j1" Gain="6" AudioFile="chan5.sph"/>
<Channel Name="chan6" Mic="c2" Transmission="u1" Gain="8" AudioFile="chan6.sph"/>
<Channel Name="chan7" Mic="c2" Transmission="u1" Gain="8" AudioFile="chan7.sph"/>
<Channel Name="chan8" Mic="p1" Transmission="j1" Gain="13" AudioFile="chan8.sph"/>
<Channel Name="chanB" Mic="u1" Transmission="j1" Gain="11" 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="8" AudioFile="chanE.sph"/>
<Channel Name="chanF" Mic="c2" Transmission="u1" Gain="8" AudioFile="chanF.sph"/>
</Channels>
<Participants DB="icsi1.spk">
<Participant Name="mn014" Channel="chan1" Seat="2"/>
<Participant Name="fe008" Channel="chan2" Seat="1"/>
<Participant Name="me013" Channel="chan3" Seat="6"/>
<Participant Name="me018" Channel="chan4" Seat="7"/>
<Participant Name="me001" Channel="chan5" Seat="3"/>
<Participant Name="me011" Channel="chan8" Seat="8"/>
<Participant Name="fe016" Channel="chanB" Seat="9"/>
</Participants>
</Preamble>
<Transcript StartTime="0.0" EndTime="2907.472">
<Segment StartTime="0.000" EndTime="7.540" CloseMic="false">
<Comment Description="Meeting starts 1 min. 38 secs into recording. Due to chaos of people checking mikes, et cetera. Transcription starts also at 1 min. 38 sec. Presegs prior to that are left available, though, just in case &quot;pre-meeting&quot; becomes of interest."/>
</Segment>
<Segment StartTime="0.620" EndTime="1.520" Participant="me013">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="0.750" EndTime="1.710" Participant="fe008">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="1.910" EndTime="3.050" Participant="fe008">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="2.620" EndTime="3.540" Participant="me001">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="3.060" EndTime="4.800" Participant="fe008">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="3.600" EndTime="4.760" Participant="mn014">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="5.110" EndTime="7.280" Participant="me001">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="5.200" EndTime="6.620" Participant="me018">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="5.360" EndTime="7.160" Participant="mn014">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="10.750" EndTime="11.530" Participant="fe008">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="11.030" EndTime="11.530" Participant="mn014">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="11.530" EndTime="14.590" Participant="me013">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="14.080" EndTime="15.650" Participant="me001">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="14.620" EndTime="15.930" Participant="me018">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="16.390" EndTime="18.170" Participant="me013">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="17.940" EndTime="18.610" Participant="me001">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="18.170" EndTime="18.370" Participant="me018">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="18.430" EndTime="19.310" Participant="me013">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="19.310" EndTime="23.620" Participant="me018">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="20.090" EndTime="22.950" Participant="mn014">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="20.340" EndTime="23.620" Participant="fe008">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="22.090" EndTime="23.290" Participant="me001">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="23.640" EndTime="28.330" Participant="fe008">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="23.690" EndTime="28.040" Participant="me013">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="25.100" EndTime="27.560" Participant="me018">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="28.520" EndTime="30.190" Participant="me013">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="28.770" EndTime="30.290" Participant="me001">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="29.380" EndTime="30.190" Participant="fe008">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="30.190" EndTime="30.800" Participant="me018">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="30.800" EndTime="32.100" Participant="mn014">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="30.800" EndTime="32.880" Participant="fe008">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="30.800" EndTime="33.220" Participant="me013">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="32.530" EndTime="33.480" Participant="me001">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="33.680" EndTime="35.090" Participant="me018">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="33.910" EndTime="35.090" Participant="me013">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="35.640" EndTime="36.770" Participant="me001">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="37.930" EndTime="46.130" Participant="fe008">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="38.080" EndTime="39.690" Participant="me018">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="39.320" EndTime="40.180" Participant="me001">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="39.990" EndTime="41.260" Participant="me013">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="41.160" EndTime="42.040" Participant="me018">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="46.130" EndTime="48.570" Participant="me018">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="47.140" EndTime="48.940" Participant="fe008">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="48.570" EndTime="49.330" Participant="me013">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="49.890" EndTime="51.790" Participant="me013">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="50.570" EndTime="52.850" Participant="fe008">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="53.770" EndTime="57.790" Participant="me013">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="53.770" EndTime="54.690" Participant="fe008">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="53.800" EndTime="54.440" Participant="me001">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="53.840" EndTime="54.520" Participant="mn014">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="54.440" EndTime="56.320" Participant="me011">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="57.790" EndTime="62.450" Participant="me011">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="59.100" EndTime="60.050" Participant="fe008">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="62.820" EndTime="63.600" Participant="me011">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="62.900" EndTime="63.730" Participant="fe008">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="64.920" EndTime="67.740" Participant="me011">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="65.660" EndTime="67.930" Participant="me018">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="67.930" EndTime="74.250" Participant="me011">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="73.020" EndTime="73.030" Participant="me013">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="73.570" EndTime="74.900" Participant="me018">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="74.640" EndTime="74.900" Participant="me011">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="75.100" EndTime="78.680" Participant="me011">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="78.930" EndTime="84.930" Participant="fe016">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="84.930" EndTime="89.820" Participant="me011">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="86.580" EndTime="89.050" Participant="fe016">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="89.250" EndTime="89.840" Participant="me013">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="90.010" EndTime="90.070" Participant="me013">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="90.130" EndTime="98.018" Participant="me013">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="90.860" EndTime="91.970" Participant="me011">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="92.020" EndTime="92.570" Participant="me018">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="92.120" EndTime="95.420" Participant="fe008">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="92.495" EndTime="93.033" Participant="fe016">
Oh.
</Segment>
<Segment StartTime="93.460" EndTime="93.700" Participant="me018">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="95.780" EndTime="96.530" Participant="fe008">
<VocalSound Description="pre-meeting utterance"/>
</Segment>
<Segment StartTime="96.127" EndTime="97.123" Participant="me011">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="98.018" EndTime="99.904" Participant="me013">
So, uh -
</Segment>
<Segment StartTime="98.190" EndTime="98.830" Participant="fe008">
Alright.
</Segment>
<Segment StartTime="98.717" EndTime="101.008" Participant="me011">
Um, so I wanted to discuss digits briefly,
</Segment>
<Segment StartTime="99.904" EndTime="101.040" Participant="me013">
Oh good.
</Segment>
<Segment StartTime="101.008" EndTime="102.383" Participant="me011">
but that won't take too long.
</Segment>
<Segment StartTime="101.040" EndTime="101.510" Participant="me013">
Right.
</Segment>
<Segment StartTime="102.300" EndTime="104.310" Participant="me013">
O_K, agenda items,
</Segment>
<Segment StartTime="104.750" EndTime="106.241" Participant="me013">
Uh, we have digits,
</Segment>
<Segment StartTime="105.608" EndTime="106.583" Participant="me011">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="106.241" EndTime="107.192" Participant="me013">
What else we got?
</Segment>
<Segment StartTime="107.700" EndTime="109.812" Participant="mn014">
New version of the presegmentation.
</Segment>
<Segment StartTime="108.465" EndTime="109.312" Participant="me011">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="109.720" EndTime="112.540" Participant="me013">
New version of presegmentation.
</Segment>
<Segment StartTime="112.768" EndTime="114.203" Participant="me011">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="113.355" EndTime="114.160" Participant="fe008">
Um,
</Segment>
<Segment StartTime="114.610" EndTime="116.035" Participant="fe008">
do we wanna say something about the,
</Segment>
<Segment StartTime="115.310" EndTime="117.330" Participant="fe016">
Yeah, why don't you summarize the -
</Segment>
<Segment StartTime="116.373" EndTime="119.282" Participant="fe008">
an update of the, uh, transcript?
</Segment>
<Segment StartTime="119.650" EndTime="121.693" Participant="me013">
Update on transcripts.
</Segment>
<Segment StartTime="121.831" EndTime="123.544" Participant="fe016">
And I guess that includes some -
</Segment>
<Segment StartTime="124.022" EndTime="125.557" Participant="fe016">
the filtering for the,
</Segment>
<Segment StartTime="125.875" EndTime="127.589" Participant="fe016">
the A_S_I refs, too.
</Segment>
<Segment StartTime="128.278" EndTime="128.716" Participant="fe008">
Mmm.
</Segment>
<Segment StartTime="129.070" EndTime="130.360" Participant="me013">
Filtering for what?
</Segment>
<Segment StartTime="130.283" EndTime="132.917" Participant="fe016">
For the references that we need to go from the -
</Segment>
<Segment StartTime="133.216" EndTime="136.583" Participant="fe016">
the <Pause/> <Emphasis> fancy </Emphasis> transcripts to the sort of <NonVocalSound Description="mike noise"/> <Emphasis> brain-dead. </Emphasis>
</Segment>
<Segment StartTime="135.171" EndTime="140.120" Participant="fe008">
It'll - it'll be - basically it'll be a re-cap of a meeting that we had jointly this morning.
</Segment>
<Segment StartTime="137.790" EndTime="138.590" Participant="me013">
Uh-huh.
</Segment>
<Segment StartTime="139.473" EndTime="140.629" Participant="fe016">
With Don, as well.
</Segment>
<Segment StartTime="140.409" EndTime="140.907" Participant="fe008">
Mm-hmm.
</Segment>
<Segment StartTime="140.983" EndTime="141.840" Participant="me013">
Got it.
</Segment>
<Segment StartTime="143.611" EndTime="145.922" Participant="me013">
Anything else more pressing than those things? So -
</Segment>
<Segment StartTime="144.731" EndTime="145.863" Participant="me011">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="145.922" EndTime="149.230" Participant="me013">
So, why don't we just do those. You said yours was brief, so -
</Segment>
<Segment StartTime="147.640" EndTime="148.118" Participant="me011">
O_K.
</Segment>
<Segment StartTime="149.522" EndTime="150.724" Participant="me011">
O_K well, the, w- uh
</Segment>
<Segment StartTime="150.724" EndTime="153.373" Participant="me011">
as you can see from the numbers on the digits we're almost done.
</Segment>
<Segment StartTime="153.373" EndTime="156.043" Participant="me011">
The digits goes up to <Pause/> about four thousand.
</Segment>
<Segment StartTime="156.700" EndTime="157.199" Participant="me011">
Um,
</Segment>
<Segment StartTime="157.740" EndTime="158.902" Participant="me011">
and so, uh,
</Segment>
<Segment StartTime="159.473" EndTime="162.203" Participant="me011">
we probably will be done with the T_I-digits in,
</Segment>
<Segment StartTime="162.721" EndTime="164.036" Participant="me011">
um, another couple weeks.
</Segment>
<Segment StartTime="164.460" EndTime="166.466" Participant="me011">
um, depending on how many we read each time.
</Segment>
<Segment StartTime="166.910" EndTime="168.404" Participant="me011">
So there were a bunch that we skipped.
</Segment>
<Segment StartTime="168.404" EndTime="171.877" Participant="me011">
You know, someone fills out the form and then they're not at the meeting and so it's blank.
</Segment>
<Segment StartTime="172.210" EndTime="174.286" Participant="me011">
Um, but those are almost all filled in as well.
</Segment>
<Segment StartTime="174.763" EndTime="179.466" Participant="me011">
And so, once we're - it's <Emphasis> done </Emphasis> it would be very nice to train up a recognizer and actually start working with this data.
</Segment>
<Segment StartTime="179.559" EndTime="183.010" Participant="me018">
So we'll have a corpus that's the size of T_I-digits?
</Segment>
<Segment StartTime="179.764" EndTime="180.342" Participant="me011">
And so -
</Segment>
<Segment StartTime="182.868" EndTime="184.949" Participant="me011">
One particular test set of T_I-digits.
</Segment>
<Segment StartTime="185.030" EndTime="185.853" Participant="me018">
Test set, O_K.
</Segment>
<Segment StartTime="185.120" EndTime="186.813" Participant="me011">
So, I - I extracted,
</Segment>
<Segment StartTime="187.710" EndTime="191.602" Participant="me011">
Ther- there was a file sitting around which people have used here as a test set.
</Segment>
<Segment StartTime="191.784" EndTime="195.610" Participant="me011">
It had been randomized and so on and that's just what I used to generate the order.
</Segment>
<Segment StartTime="193.334" EndTime="194.957" Participant="me018">
Oh! Great.
</Segment>
<Segment StartTime="195.849" EndTime="197.164" Participant="me011">
of these particular ones.
</Segment>
<Segment StartTime="195.961" EndTime="196.406" Participant="me018">
Great.
</Segment>
<Segment StartTime="197.690" EndTime="198.100" Participant="me011">
Um -
</Segment>
<Segment StartTime="198.032" EndTime="199.610" Participant="me013">
So, I'm impressed by what we could do,
</Segment>
<Segment StartTime="199.610" EndTime="204.600" Participant="me013">
Is take the standard training set for T_I-digits, train up with whatever, you know,
</Segment>
<Segment StartTime="204.600" EndTime="207.427" Participant="me013">
great features we think we have, uh for instance,
</Segment>
<Segment StartTime="207.427" EndTime="209.946" Participant="me013">
and then test on uh this test set.
</Segment>
<Segment StartTime="209.946" EndTime="218.300" Participant="me013">
And presumably uh it should do reasonably well on that, and then, presumably, we should go to the distant mike, and it should do poorly.
</Segment>
<Segment StartTime="214.805" EndTime="215.307" Participant="me018">
Yeah.
</Segment>
<Segment StartTime="217.910" EndTime="218.438" Participant="me011">
Right.
</Segment>
<Segment StartTime="218.360" EndTime="222.317" Participant="me013">
And then we should get really smart over the next year or two, and it - that should get better.
</Segment>
<Segment StartTime="218.438" EndTime="219.054" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="221.252" EndTime="224.211" Participant="me011">
And inc- increase it by one or two percent, yeah. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="222.317" EndTime="222.782" Participant="me013">
Yeah,
</Segment>
<Segment StartTime="222.782" EndTime="224.707" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="223.282" EndTime="224.272" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="223.822" EndTime="225.850" Participant="me018">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="224.510" EndTime="225.307" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="224.707" EndTime="225.440" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="225.307" EndTime="226.163" Participant="me011">
Um, but,
</Segment>
<Segment StartTime="226.524" EndTime="229.425" Participant="me011">
in order to do that we need to extract out the actual digits.
</Segment>
<Segment StartTime="229.596" EndTime="230.020" Participant="me013">
Right.
</Segment>
<Segment StartTime="229.931" EndTime="230.623" Participant="me011">
Um,
</Segment>
<Segment StartTime="230.623" EndTime="237.439" Participant="me011">
so that - the reason it's not just a transcript is that there're false starts, and misreads, and miscues and things like that.
</Segment>
<Segment StartTime="237.439" EndTime="240.766" Participant="me011">
And so I have a set of scripts and X_Waves where you just select the portion,
</Segment>
<Segment StartTime="240.766" EndTime="241.563" Participant="me011">
hit R_,
</Segment>
<Segment StartTime="241.962" EndTime="244.264" Participant="me011">
um, it tells you what the next one should be,
</Segment>
<Segment StartTime="244.264" EndTime="245.320" Participant="me011">
and you just look for that.
</Segment>
<Segment StartTime="245.320" EndTime="247.063" Participant="me011">
You know, so it - it'll put on the screen,
</Segment>
<Segment StartTime="247.063" EndTime="248.388" Participant="me011">
"The next set is six nine,
</Segment>
<Segment StartTime="248.687" EndTime="249.643" Participant="me011">
nine two two".
</Segment>
<Segment StartTime="250.059" EndTime="250.936" Participant="me011">
And you find that,
</Segment>
<Segment StartTime="250.936" EndTime="251.454" Participant="me011">
and,
</Segment>
<Segment StartTime="251.454" EndTime="254.184" Participant="me011">
hit the key and it records it in a file in a particular format.
</Segment>
<Segment StartTime="254.430" EndTime="255.076" Participant="me013">
So is this -
</Segment>
<Segment StartTime="254.463" EndTime="254.961" Participant="me011">
And so the -
</Segment>
<Segment StartTime="254.961" EndTime="258.377" Participant="me011">
the <Emphasis> question </Emphasis> is, should we have the transcribers do that or should we just do it?
</Segment>
<Segment StartTime="258.786" EndTime="260.883" Participant="me011">
Well, some of us. I've been do- I've done,
</Segment>
<Segment StartTime="260.883" EndTime="262.517" Participant="me011">
eight meetings, something like that,
</Segment>
<Segment StartTime="262.517" EndTime="263.434" Participant="me011">
just by hand.
</Segment>
<Segment StartTime="264.349" EndTime="265.504" Participant="me011">
Just myself, rather.
</Segment>
<Segment StartTime="267.277" EndTime="268.752" Participant="me011">
So it will not take long.
</Segment>
<Segment StartTime="268.752" EndTime="269.464" Participant="me011">
Um -
</Segment>
<Segment StartTime="270.590" EndTime="271.928" Participant="me013">
Uh, what - what do you think?
</Segment>
<Segment StartTime="271.238" EndTime="280.111" Participant="fe008">
My feeling is that- we discussed this right before coffee and I think it's a - it's a fine idea partly because, um, it's not un- unrelated to their present skill set,
</Segment>
<Segment StartTime="280.443" EndTime="286.759" Participant="fe008">
but it will add, for them, an extra dimension, it might be an interesting break for them. And also it is contributing to the, uh,
</Segment>
<Segment StartTime="286.759" EndTime="293.832" Participant="fe008">
c- composition of the transcript cuz we can incorporate those numbers directly and it'll be a more complete transcript. So I'm - I think it's fine, that part.
</Segment>
<Segment StartTime="293.366" EndTime="294.343" Participant="me011">
There is - there is -
</Segment>
<Segment StartTime="293.950" EndTime="296.042" Participant="me013">
So you think it's fine to have the <Emphasis> transcribers </Emphasis> do it?
</Segment>
<Segment StartTime="295.765" EndTime="296.303" Participant="fe008">
Mm-hmm.
</Segment>
<Segment StartTime="296.042" EndTime="296.985" Participant="me013">
Yeah, O_K.
</Segment>
<Segment StartTime="296.472" EndTime="301.533" Participant="me011">
There's one other small bit, which is just entering the information which at s- which is at the top of this form,
</Segment>
<Segment StartTime="297.349" EndTime="297.668" Participant="me018">
Good.
</Segment>
<Segment StartTime="301.533" EndTime="302.549" Participant="me011">
onto the computer,
</Segment>
<Segment StartTime="302.848" EndTime="305.837" Participant="me011">
to go along with the - where the digits are recorded automatically.
</Segment>
<Segment StartTime="303.524" EndTime="304.420" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="306.116" EndTime="306.833" Participant="me011">
And so it's just,
</Segment>
<Segment StartTime="306.833" EndTime="309.881" Participant="me011">
you know, typing in name, times - time, date, and so on.
</Segment>
<Segment StartTime="310.332" EndTime="312.398" Participant="me011">
Um, which again either <Emphasis> they </Emphasis> can do,
</Segment>
<Segment StartTime="312.597" EndTime="313.593" Participant="me011">
but it is, you know,
</Segment>
<Segment StartTime="313.952" EndTime="315.084" Participant="me011">
firing up an editor,
</Segment>
<Segment StartTime="314.339" EndTime="316.471" Participant="me018">
<VocalSound Description="swallow"/>
</Segment>
<Segment StartTime="315.641" EndTime="316.228" Participant="me011">
or,
</Segment>
<Segment StartTime="316.228" EndTime="317.450" Participant="me011">
again, I can do.
</Segment>
<Segment StartTime="317.802" EndTime="318.833" Participant="me011">
Or someone else can do.
</Segment>
<Segment StartTime="318.581" EndTime="320.414" Participant="fe008">
And, that, you know, I'm not,
</Segment>
<Segment StartTime="320.892" EndTime="323.741" Participant="fe008">
that - that one I'm not so sure if it's into the - the,
</Segment>
<Segment StartTime="324.200" EndTime="324.886" Participant="fe008">
things that,
</Segment>
<Segment StartTime="324.886" EndTime="325.524" Participant="fe008">
I,
</Segment>
<Segment StartTime="325.524" EndTime="327.496" Participant="fe008">
wanted to use the hours for, because the,
</Segment>
<Segment StartTime="327.496" EndTime="330.804" Participant="fe008">
the time that they'd be spending doing that they wouldn't be able to be putting more words on.
</Segment>
<Segment StartTime="330.981" EndTime="331.645" Participant="me013">
Mmm.
</Segment>
<Segment StartTime="331.581" EndTime="333.060" Participant="fe008">
But that's really your choice, it's your -
</Segment>
<Segment StartTime="332.571" EndTime="334.978" Participant="me018">
So are these two <Emphasis> separate </Emphasis> tasks that can happen?
</Segment>
<Segment StartTime="334.978" EndTime="336.829" Participant="me018">
Or do they have to happen at the same time before -
</Segment>
<Segment StartTime="336.302" EndTime="337.304" Participant="me011">
No they don't have - this -
</Segment>
<Segment StartTime="337.760" EndTime="339.305" Participant="me011">
you have to enter the data <Emphasis> before, </Emphasis>
</Segment>
<Segment StartTime="339.305" EndTime="342.373" Participant="me011">
you do the second task, but they don't have to happen at the same time.
</Segment>
<Segment StartTime="343.629" EndTime="346.904" Participant="me011">
So it's - it's just I have a file whi- which has this information on it,
</Segment>
<Segment StartTime="343.740" EndTime="344.097" Participant="me018">
O_K.
</Segment>
<Segment StartTime="346.904" EndTime="349.056" Participant="me011">
and then when you start using my scripts,
</Segment>
<Segment StartTime="349.056" EndTime="350.610" Participant="me011">
for extracting the times,
</Segment>
<Segment StartTime="350.610" EndTime="352.384" Participant="me011">
it adds the times at the bottom of the file.
</Segment>
<Segment StartTime="353.738" EndTime="354.912" Participant="me011">
And so, um,
</Segment>
<Segment StartTime="354.912" EndTime="359.163" Participant="me011">
I mean, it's easy to create the files and leave them <Emphasis> blank, </Emphasis> and so actually we could do it in either order.
</Segment>
<Segment StartTime="359.019" EndTime="359.593" Participant="me018">
Oh, O_K.
</Segment>
<Segment StartTime="359.921" EndTime="360.498" Participant="me011">
Um,
</Segment>
<Segment StartTime="360.817" EndTime="364.194" Participant="me011">
it's - it's <Emphasis> sort </Emphasis> of nice to have the same person do it just as a double-check,
</Segment>
<Segment StartTime="364.493" EndTime="366.445" Participant="me011">
to make sure you're entering for the right person.
</Segment>
<Segment StartTime="366.445" EndTime="366.923" Participant="me011">
But,
</Segment>
<Segment StartTime="368.278" EndTime="368.823" Participant="me011">
either way.
</Segment>
<Segment StartTime="369.287" EndTime="369.978" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="370.620" EndTime="373.870" Participant="me013">
Yeah just by way of uh, uh, a uh,
</Segment>
<Segment StartTime="374.312" EndTime="374.967" Participant="me011">
<VocalSound Description="clear throat"/>
</Segment>
<Segment StartTime="374.780" EndTime="376.410" Participant="me013">
order of magnitude, uh,
</Segment>
<Segment StartTime="376.810" EndTime="380.460" Participant="me013">
um, we've been working with this Aurora, uh data set.
</Segment>
<Segment StartTime="380.530" EndTime="383.050" Participant="me013">
And, uh, the best score,
</Segment>
<Segment StartTime="381.691" EndTime="382.507" Participant="me018">
<VocalSound Description="yawn"/>
</Segment>
<Segment StartTime="383.050" EndTime="384.320" Participant="me013">
on the,
</Segment>
<Segment StartTime="384.320" EndTime="391.683" Participant="me013">
<Emphasis> nicest </Emphasis> part of the data, that is, where you've got training and test set that are basically the same kinds of noise and so forth,
</Segment>
<Segment StartTime="391.683" EndTime="394.221" Participant="me013">
uh, is about, uh -
</Segment>
<Segment StartTime="394.221" EndTime="397.360" Participant="me013">
I think the best score was something like five percent,
</Segment>
<Segment StartTime="397.499" EndTime="399.910" Participant="me013">
uh, error, per digit.
</Segment>
<Segment StartTime="400.340" EndTime="401.470" Participant="me013">
So, that -
</Segment>
<Segment StartTime="400.533" EndTime="401.493" Participant="me011">
Per digit.
</Segment>
<Segment StartTime="400.708" EndTime="401.764" Participant="mn014">
Per digit. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="401.530" EndTime="404.359" Participant="me013">
You're right. So if you were doing <Pause/> ten digit,
</Segment>
<Segment StartTime="404.780" EndTime="408.857" Participant="me013">
uh, recognition, <VocalSound Description="laugh"/> you would really be in trouble.
</Segment>
<Segment StartTime="408.733" EndTime="409.194" Participant="me018">
Mm-hmm.
</Segment>
<Segment StartTime="408.857" EndTime="410.526" Participant="me013">
So - So the -
</Segment>
<Segment StartTime="410.526" EndTime="415.310" Participant="me013">
The point there, and this is uh car noise uh, uh things, but - but real -
</Segment>
<Segment StartTime="415.310" EndTime="417.390" Participant="me013">
real situation, well, "real",
</Segment>
<Segment StartTime="417.450" EndTime="424.550" Participant="me013">
Um, the - uh there's one microphone that's close, that they have as - as <Emphasis> this </Emphasis> sort of thing, close versus distant.
</Segment>
<Segment StartTime="424.811" EndTime="429.591" Participant="me013">
Uh but in a <Emphasis> car, </Emphasis> instead of - instead of having a projector noise it's - it's car noise.
</Segment>
<Segment StartTime="428.505" EndTime="428.884" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="429.909" EndTime="436.050" Participant="me013">
Uh but it wasn't artificially added to get some - some artificial signal-to-noise ratio. It was just people driving around in a car.
</Segment>
<Segment StartTime="431.874" EndTime="432.791" Participant="me018">
<VocalSound Description="swallow"/>
</Segment>
<Segment StartTime="436.250" EndTime="439.074" Participant="me013">
So, that's - that's an indication, uh that was with,
</Segment>
<Segment StartTime="439.371" EndTime="443.364" Participant="me013">
many sites competing, and this was the very <Emphasis> best </Emphasis> score and so forth, so.
</Segment>
<Segment StartTime="443.380" EndTime="445.506" Participant="me018">
Although the models weren't,
</Segment>
<Segment StartTime="443.775" EndTime="445.277" Participant="me013">
More typical numbers like
</Segment>
<Segment StartTime="445.905" EndTime="447.278" Participant="me018">
that good, right? I mean,
</Segment>
<Segment StartTime="447.278" EndTime="448.604" Participant="me018">
the models are pretty crappy?
</Segment>
<Segment StartTime="448.890" EndTime="454.909" Participant="me013">
You're right. I think that we could have done better on the models, but the thing is that we got - this - this is the kind of typical number,
</Segment>
<Segment StartTime="455.474" EndTime="457.949" Participant="me013">
for all of the, uh, uh,
</Segment>
<Segment StartTime="458.093" EndTime="460.535" Participant="me013">
things in this task, all of the, um,
</Segment>
<Segment StartTime="460.640" EndTime="461.383" Participant="me013">
languages.
</Segment>
<Segment StartTime="461.636" EndTime="465.750" Participant="me013">
And so I - I think we'd probably - the models would be better in some than in others. Um,
</Segment>
<Segment StartTime="465.017" EndTime="465.376" Participant="me018">
Hmm.
</Segment>
<Segment StartTime="466.287" EndTime="467.500" Participant="me013">
so, uh.
</Segment>
<Segment StartTime="467.500" EndTime="472.380" Participant="me013">
<Emphasis> Anyway, </Emphasis> just an indication once you get into this kind of realm even if you're looking at connected digits it can be pretty hard.
</Segment>
<Segment StartTime="474.094" EndTime="474.512" Participant="fe008">
Hmm.
</Segment>
<Segment StartTime="475.309" EndTime="476.923" Participant="fe008">
It's gonna be fun to see how we,
</Segment>
<Segment StartTime="477.182" EndTime="477.979" Participant="fe008">
compare at this.
</Segment>
<Segment StartTime="477.984" EndTime="479.842" Participant="me018">
<VocalSound Description="breath"/> How did we do on the T_I-digits?
</Segment>
<Segment StartTime="478.059" EndTime="478.455" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="478.278" EndTime="478.975" Participant="fe008">
Very exciting.
</Segment>
<Segment StartTime="478.490" EndTime="481.093" Participant="me011">
Well the <Emphasis> prosodics </Emphasis> are so much different s- it's gonna be,
</Segment>
<Segment StartTime="481.551" EndTime="482.388" Participant="me011">
strange.
</Segment>
<Segment StartTime="482.952" EndTime="485.913" Participant="me011">
I mean the prosodics are not the same as T_I-digits, for example.
</Segment>
<Segment StartTime="486.050" EndTime="486.344" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="486.511" EndTime="488.873" Participant="me011">
So I'm - I'm not sure how much of effect that will have.
</Segment>
<Segment StartTime="488.657" EndTime="489.395" Participant="me018">
H- how do -
</Segment>
<Segment StartTime="488.677" EndTime="490.321" Participant="fe016">
What do you mean, the prosodics?
</Segment>
<Segment StartTime="490.288" EndTime="492.453" Participant="me011">
Um, just what we were talking about with grouping.
</Segment>
<Segment StartTime="492.872" EndTime="494.325" Participant="me011">
That with these, the grouping,
</Segment>
<Segment StartTime="494.325" EndTime="496.158" Participant="me011">
there's no grouping at all, and so it's just -
</Segment>
<Segment StartTime="496.158" EndTime="497.573" Participant="me011">
the only sort of
</Segment>
<Segment StartTime="498.111" EndTime="500.104" Participant="me011">
discontinuity you have is at the beginning and the end.
</Segment>
<Segment StartTime="500.342" EndTime="504.605" Participant="fe016">
So what are they doing in Aurora, are they reading actual phone numbers, or,
</Segment>
<Segment StartTime="502.662" EndTime="503.850" Participant="me011">
Aurora I don't know.
</Segment>
<Segment StartTime="504.089" EndTime="505.543" Participant="me011">
I don't know what they do in Aurora.
</Segment>
<Segment StartTime="504.605" EndTime="506.338" Participant="fe016">
a - a digit at a time, or - ? <Comment Description="leading question"/>
</Segment>
<Segment StartTime="506.535" EndTime="510.105" Participant="me013">
Uh, I'm not sure how - no, no I mean it's connected - it's connected, uh,
</Segment>
<Segment StartTime="506.836" EndTime="507.434" Participant="fe016">
Cuz it's -
</Segment>
<Segment StartTime="509.550" EndTime="510.449" Participant="fe016">
Connected.
</Segment>
<Segment StartTime="510.366" EndTime="511.169" Participant="me013">
digits, yeah.
</Segment>
<Segment StartTime="511.472" EndTime="512.539" Participant="fe016">
So there's also the -
</Segment>
<Segment StartTime="511.520" EndTime="511.790" Participant="me013">
But.
</Segment>
<Segment StartTime="511.572" EndTime="511.891" Participant="me011">
But -
</Segment>
<Segment StartTime="512.539" EndTime="514.372" Participant="fe016">
not just the prosody but the cross -
</Segment>
<Segment StartTime="514.372" EndTime="516.782" Participant="fe016">
the cross-word modeling is probably quite different.
</Segment>
<Segment StartTime="515.377" EndTime="515.841" Participant="me011">
Right.
</Segment>
<Segment StartTime="516.832" EndTime="518.486" Participant="me011">
But in T_I-digits,
</Segment>
<Segment StartTime="516.873" EndTime="517.260" Participant="me018">
H-
</Segment>
<Segment StartTime="518.486" EndTime="522.809" Participant="me011">
they're reading things <Emphasis> like </Emphasis> zip codes and phone numbers and things like that, so it's gonna be different.
</Segment>
<Segment StartTime="521.172" EndTime="521.607" Participant="fe016">
Right.
</Segment>
<Segment StartTime="521.832" EndTime="523.514" Participant="me018">
How do we do on T_I-digits?
</Segment>
<Segment StartTime="524.270" EndTime="526.143" Participant="me011">
I don't remember. I mean, very good, right?
</Segment>
<Segment StartTime="526.270" EndTime="528.260" Participant="me013">
Yeah, I mean we were in the.
</Segment>
<Segment StartTime="527.438" EndTime="529.703" Participant="me011">
One and a half percent, two percent, something like that?
</Segment>
<Segment StartTime="528.920" EndTime="532.463" Participant="me013">
Uh, I th- no I think we got under a percent, but it was - but it's - but I mean.
</Segment>
<Segment StartTime="531.078" EndTime="531.994" Participant="me011">
Oh really? O_K.
</Segment>
<Segment StartTime="532.002" EndTime="532.520" Participant="fe008">
s-
</Segment>
<Segment StartTime="532.652" EndTime="539.475" Participant="me013">
The <Emphasis> very </Emphasis> best system that I saw in the literature was a point two five percent or something that somebody had at - at Bell Labs, or.
</Segment>
<Segment StartTime="539.382" EndTime="539.873" Participant="me011">
Alright.
</Segment>
<Segment StartTime="539.850" EndTime="540.562" Participant="me013">
Uh, but.
</Segment>
<Segment StartTime="539.861" EndTime="540.459" Participant="me018">
Hmm.
</Segment>
<Segment StartTime="540.949" EndTime="547.762" Participant="me013">
But, uh, sort of pulling out all the stops. But I think a <Emphasis> lot </Emphasis> of systems sort of get half a percent, or three-quarters a percent, and we're - we're in there somewhere.
</Segment>
<Segment StartTime="545.892" EndTime="546.291" Participant="me011">
Right.
</Segment>
<Segment StartTime="547.502" EndTime="551.978" Participant="me011">
But that - I mean it's really - it's - it's close-talking mikes, no noise, clean signal,
</Segment>
<Segment StartTime="550.466" EndTime="551.063" Participant="fe016">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="551.978" EndTime="553.359" Participant="me011">
just digits, I mean,
</Segment>
<Segment StartTime="553.159" EndTime="553.555" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="553.618" EndTime="554.853" Participant="me011">
every- everything is good.
</Segment>
<Segment StartTime="553.784" EndTime="556.443" Participant="fe016">
It's the beginning of time in speech recognition.
</Segment>
<Segment StartTime="555.212" EndTime="556.188" Participant="me011">
Yes, exactly.
</Segment>
<Segment StartTime="555.938" EndTime="558.139" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="555.969" EndTime="558.193" Participant="me018">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="558.139" EndTime="558.605" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="558.870" EndTime="561.546" Participant="me011">
And we've only recently got it to anywhere near human.
</Segment>
<Segment StartTime="559.867" EndTime="560.718" Participant="fe016">
It's like the,
</Segment>
<Segment StartTime="560.718" EndTime="562.093" Participant="fe016">
single cell,
</Segment>
<Segment StartTime="560.892" EndTime="561.348" Participant="me018">
Pre-
</Segment>
<Segment StartTime="562.093" EndTime="562.770" Participant="fe016">
you know,
</Segment>
<Segment StartTime="562.125" EndTime="563.739" Participant="me018">
prehistory.<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="563.069" EndTime="564.584" Participant="fe016">
it's the beginning of life, yeah.
</Segment>
<Segment StartTime="564.894" EndTime="565.565" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="564.933" EndTime="567.974" Participant="me011">
And it's still like an order of magnitude worse than what humans do. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="567.711" EndTime="568.250" Participant="fe016">
Right.
</Segment>
<Segment StartTime="568.346" EndTime="568.692" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="568.545" EndTime="569.143" Participant="me011">
So.
</Segment>
<Segment StartTime="569.047" EndTime="570.476" Participant="me013">
When - When they're wide awake, yeah.
</Segment>
<Segment StartTime="570.381" EndTime="571.337" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="570.942" EndTime="571.459" Participant="me011">
Yeah.
</Segment>
<Segment StartTime="571.148" EndTime="573.920" Participant="me013">
Um, <VocalSound Description="laugh"/> after coffee, you're right.
</Segment>
<Segment StartTime="571.211" EndTime="571.847" Participant="me018">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="571.257" EndTime="572.006" Participant="fe016">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="571.459" EndTime="572.455" Participant="me011">
After coffee.
</Segment>
<Segment StartTime="572.623" EndTime="573.041" Participant="me018">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="572.955" EndTime="573.414" Participant="mn014">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="574.238" EndTime="575.230" Participant="me013">
Not after lunch.
</Segment>
<Segment StartTime="575.064" EndTime="576.312" Participant="me011">
O_K, so, um,
</Segment>
<Segment StartTime="576.618" EndTime="578.832" Participant="me011">
what I'll do then is I'll go ahead and enter,
</Segment>
<Segment StartTime="579.687" EndTime="580.384" Participant="me011">
this data.
</Segment>
<Segment StartTime="581.041" EndTime="582.835" Participant="me011">
And then, hand off to Jane,
</Segment>
<Segment StartTime="583.412" EndTime="586.431" Participant="me011">
and the transcribers to do the actual extraction of the digits.
</Segment>
<Segment StartTime="584.705" EndTime="585.025" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="586.735" EndTime="587.040" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="587.560" EndTime="588.519" Participant="me013">
One question I have that -
</Segment>
<Segment StartTime="588.642" EndTime="590.860" Participant="me013">
that- I mean, we wouldn't know the answer to now but might,
</Segment>
<Segment StartTime="590.400" EndTime="590.679" Participant="me011">
Hmm.
</Segment>
<Segment StartTime="591.242" EndTime="594.632" Participant="me013">
do some guessing, but I was talking before about doing some model- modeling of arti- uh,
</Segment>
<Segment StartTime="594.977" EndTime="597.090" Participant="me013">
uh, <Emphasis> marking </Emphasis> of articulatory,
</Segment>
<Segment StartTime="597.230" EndTime="597.930" Participant="me013">
features,
</Segment>
<Segment StartTime="598.332" EndTime="599.475" Participant="me013">
with overlap and so on.
</Segment>
<Segment StartTime="599.662" EndTime="600.193" Participant="me013">
And,
</Segment>
<Segment StartTime="600.193" EndTime="600.912" Participant="me013">
and,
</Segment>
<Segment StartTime="601.006" EndTime="601.444" Participant="me013">
um,
</Segment>
<Segment StartTime="602.098" EndTime="603.022" Participant="me013">
On some subset.
</Segment>
<Segment StartTime="603.470" EndTime="609.147" Participant="me013">
One thought might be to do this uh, on - on the digits, or some piece of the digits. Uh, it'd be easier,
</Segment>
<Segment StartTime="609.780" EndTime="613.405" Participant="me013">
uh, and so forth. The only thing is I'm a little concerned that maybe the <Emphasis> kind </Emphasis> of phenomena,
</Segment>
<Segment StartTime="613.787" EndTime="614.760" Participant="me013">
in w- i- i-
</Segment>
<Segment StartTime="614.760" EndTime="621.953" Participant="me013">
The reason for doing it is because the - the argument is that certainly with conversational speech, the stuff that we've looked at here before,
</Segment>
<Segment StartTime="622.629" EndTime="624.758" Participant="me013">
um, just doing the simple mapping,
</Segment>
<Segment StartTime="625.281" EndTime="626.410" Participant="me013">
from, um,
</Segment>
<Segment StartTime="626.790" EndTime="627.531" Participant="me013">
the phone,
</Segment>
<Segment StartTime="628.117" EndTime="630.784" Participant="me013">
to the corresponding features that you could look up in a book,
</Segment>
<Segment StartTime="631.371" EndTime="632.393" Participant="me013">
uh, isn't right.
</Segment>
<Segment StartTime="632.786" EndTime="637.625" Participant="me013">
It isn't actually right. In fact there's these overlapping <Pronounce> processes </Pronounce> where some voicing some up and then some, you know,
</Segment>
<Segment StartTime="637.738" EndTime="643.353" Participant="me013">
some nasality is - comes in here, and so forth. And you do this gross thing saying "Well I guess it's this phone starting there".
</Segment>
<Segment StartTime="643.988" EndTime="644.581" Participant="me013">
So,
</Segment>
<Segment StartTime="644.720" EndTime="646.140" Participant="me013">
uh, that's the reasoning. <Emphasis> But, </Emphasis>
</Segment>
<Segment StartTime="646.582" EndTime="650.632" Participant="me013">
It could be that when we're reading digits, because it's - it's for such a limited set,
</Segment>
<Segment StartTime="650.794" EndTime="651.577" Participant="me013">
that maybe -
</Segment>
<Segment StartTime="651.641" EndTime="653.706" Participant="me013">
maybe that phenomenon doesn't occur as much. I don't know.
</Segment>
<Segment StartTime="654.060" EndTime="657.640" Participant="me013">
Di- an- anybody - ? <Pause/> Do you have any - ? <Pause/> Anybody have any opinion about that, or - ? <Comment Description="leading question"/>
</Segment>
<Segment StartTime="654.147" EndTime="654.466" Participant="fe008">
<Uncertain NumSyllables="1"> @@ </Uncertain>.
</Segment>
<Segment StartTime="656.697" EndTime="661.402" Participant="fe008">
It s- strikes me that there are more - each of them is more informative because it's so,
</Segment>
<Segment StartTime="658.091" EndTime="659.366" Participant="me018">
<VocalSound Description="swallow"/>
</Segment>
<Segment StartTime="661.402" EndTime="666.483" Participant="fe008">
random, and that people might articulate more, and you that might end up with more - a closer correspondence.
</Segment>
<Segment StartTime="664.512" EndTime="664.915" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="666.621" EndTime="669.520" Participant="me011">
Yeah - that's - I - I agree. That - it's just -
</Segment>
<Segment StartTime="666.659" EndTime="666.987" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="668.350" EndTime="669.850" Participant="me018">
Sort of less predictability, and -
</Segment>
<Segment StartTime="669.668" EndTime="670.027" Participant="fe008">
Mm-hmm.
</Segment>
<Segment StartTime="669.808" EndTime="670.157" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="670.460" EndTime="672.692" Participant="fe016">
Well it's definitely true that, when people are,
</Segment>
<Segment StartTime="670.928" EndTime="671.347" Participant="me011">
It's a -
</Segment>
<Segment StartTime="673.270" EndTime="674.127" Participant="fe016">
reading,
</Segment>
<Segment StartTime="674.425" EndTime="676.069" Participant="fe016">
even if they're re-reading what,
</Segment>
<Segment StartTime="676.331" EndTime="680.674" Participant="fe016">
they had said spontaneously, that they have very different patterns. Mitch showed that, and some,
</Segment>
<Segment StartTime="680.474" EndTime="680.742" Participant="me013">
Right.
</Segment>
<Segment StartTime="681.292" EndTime="682.789" Participant="fe016">
dissertations have shown that.
</Segment>
<Segment StartTime="683.391" EndTime="686.927" Participant="fe016">
So the fact that they're reading, first of all, whether they're reading in a room of,
</Segment>
<Segment StartTime="686.927" EndTime="690.425" Participant="fe016">
people, or rea- you know, just the fact that they're reading will make a difference.
</Segment>
<Segment StartTime="690.608" EndTime="690.978" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="691.083" EndTime="691.659" Participant="fe016">
And,
</Segment>
<Segment StartTime="691.185" EndTime="691.641" Participant="me011">
Well -
</Segment>
<Segment StartTime="691.659" EndTime="693.223" Participant="fe016">
depends what you're interested in.
</Segment>
<Segment StartTime="693.474" EndTime="694.112" Participant="me011">
Would,
</Segment>
<Segment StartTime="694.112" EndTime="696.802" Participant="me011">
this corpus really be the right one to even try that on?
</Segment>
<Segment StartTime="697.704" EndTime="698.202" Participant="fe008">
<VocalSound Description="mouth"/>
</Segment>
<Segment StartTime="697.705" EndTime="701.093" Participant="me013">
See, I don't know. So, may- maybe the thing will be do - to take some very small subset,
</Segment>
<Segment StartTime="701.183" EndTime="701.973" Participant="me013">
I- mean not have a big,
</Segment>
<Segment StartTime="702.406" EndTime="703.570" Participant="me013">
program, but take a small set,
</Segment>
<Segment StartTime="703.677" EndTime="707.554" Participant="me013">
uh, subset of the conversational speech and a small subset of the digits,
</Segment>
<Segment StartTime="708.101" EndTime="708.793" Participant="me013">
and <Pause/> look
</Segment>
<Segment StartTime="708.935" EndTime="710.850" Participant="me013">
and - and just get a feeling for it.
</Segment>
<Segment StartTime="710.953" EndTime="712.326" Participant="me013">
Um, just take a look.
</Segment>
<Segment StartTime="712.452" EndTime="712.844" Participant="me013">
Really.
</Segment>
<Segment StartTime="713.550" EndTime="714.048" Participant="fe008">
H-
</Segment>
<Segment StartTime="713.623" EndTime="715.857" Participant="me013">
Cuz I don't think anybody is, I- at least, I don't know,
</Segment>
<Segment StartTime="716.430" EndTime="717.493" Participant="me013">
of anybody, uh,
</Segment>
<Segment StartTime="718.332" EndTime="719.474" Participant="me013">
well, I don't know,
</Segment>
<Segment StartTime="719.567" EndTime="723.708" Participant="fe008">
That could - could be an interesting design, too, cuz then you'd have the com- the comparison of the,
</Segment>
<Segment StartTime="719.646" EndTime="720.331" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="719.979" EndTime="720.715" Participant="me018">
You hafta -
</Segment>
<Segment StartTime="720.331" EndTime="721.026" Participant="me013">
the <Uncertain> answers </Uncertain>.
</Segment>
<Segment StartTime="721.512" EndTime="721.910" Participant="me018">
Hey.
</Segment>
<Segment StartTime="723.708" EndTime="727.633" Participant="fe008">
uh, predictable speech versus the less predictable speech and maybe you'd find that it worked in,
</Segment>
<Segment StartTime="725.781" EndTime="726.063" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="727.633" EndTime="728.250" Participant="fe008">
in the,
</Segment>
<Segment StartTime="728.250" EndTime="729.961" Participant="fe008">
case of the pr- of the, uh, non-predictable.
</Segment>
<Segment StartTime="728.478" EndTime="728.740" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="729.298" EndTime="730.372" Participant="me018">
Hafta think about,
</Segment>
<Segment StartTime="730.372" EndTime="733.324" Participant="me018">
the particular acoustic features to mark, too, because,
</Segment>
<Segment StartTime="733.602" EndTime="733.973" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="733.902" EndTime="734.879" Participant="me018">
I mean, <Emphasis> some </Emphasis> things,
</Segment>
<Segment StartTime="734.879" EndTime="736.374" Participant="me018">
they wouldn't be able to mark, like,
</Segment>
<Segment StartTime="736.692" EndTime="737.310" Participant="me018">
uh,
</Segment>
<Segment StartTime="737.773" EndTime="739.728" Participant="me018">
you know, uh, tense lax.
</Segment>
<Segment StartTime="739.728" EndTime="740.944" Participant="me018">
Some things are really difficult.
</Segment>
<Segment StartTime="741.647" EndTime="742.177" Participant="me018">
You know,
</Segment>
<Segment StartTime="742.177" EndTime="743.114" Participant="me018">
just listening.
</Segment>
<Segment StartTime="742.330" EndTime="744.091" Participant="me011">
M- I think we can get Ohala in to,
</Segment>
<Segment StartTime="742.601" EndTime="743.019" Participant="fe008">
Well.
</Segment>
<Segment StartTime="744.370" EndTime="745.663" Participant="me011">
give us some advice on that.
</Segment>
<Segment StartTime="744.972" EndTime="745.552" Participant="me018">
Yeah.
</Segment>
<Segment StartTime="745.616" EndTime="749.103" Participant="fe008">
Also I thought you were thinking of a much more restricted set of features, that -
</Segment>
<Segment StartTime="748.988" EndTime="751.031" Participant="me013">
Yeah, but I - I - I - I was,
</Segment>
<Segment StartTime="751.902" EndTime="754.986" Participant="me013">
like he said, <VocalSound Description="laugh"/> I was gonna bring John in and ask John what he thought.
</Segment>
<Segment StartTime="752.662" EndTime="753.380" Participant="fe008">
Yeah, sure.
</Segment>
<Segment StartTime="753.977" EndTime="754.456" Participant="fe008">
Sure.
</Segment>
<Segment StartTime="754.761" EndTime="755.003" Participant="fe008">
Yeah.
</Segment>
<Segment StartTime="755.073" EndTime="755.343" Participant="me013">
Right.
</Segment>
<Segment StartTime="755.868" EndTime="759.174" Participant="me013">
But I mean you want - you want it be restrictive but you also want it to - to - to have coverage.
</Segment>
<Segment StartTime="759.089" EndTime="759.528" Participant="me011">
Right.
</Segment>
<Segment StartTime="759.575" EndTime="760.363" Participant="me013">
You know i- you should.
</Segment>
<Segment StartTime="759.770" EndTime="760.065" Participant="fe008">
Yeah <Comment Description="whispered"/>
</Segment>
<Segment StartTime="760.800" EndTime="762.428" Participant="me013">
It should be such that if you,
</Segment>
<Segment StartTime="762.534" EndTime="763.507" Participant="me013">
if you, uh,
</Segment>
<Segment StartTime="763.602" EndTime="764.792" Participant="me013">
if you had o- um,
</Segment>
<Segment StartTime="764.967" EndTime="766.098" Participant="me013">
all of the features,
</Segment>
<Segment StartTime="766.405" EndTime="769.294" Participant="me013">
determined that you - that you were uh ch- have chosen,
</Segment>
<Segment StartTime="769.650" EndTime="771.052" Participant="me013">
that that would tell you,
</Segment>
<Segment StartTime="771.417" EndTime="771.818" Participant="me013">
uh,
</Segment>
<Segment StartTime="771.934" EndTime="774.029" Participant="me013">
in the steady-state case, uh, the phone.
</Segment>
<Segment StartTime="772.729" EndTime="774.562" Participant="me018">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="774.532" EndTime="774.941" Participant="me013">
So,
</Segment>
<Segment StartTime="775.039" EndTime="775.438" Participant="fe008">
O_K.
</Segment>
<Segment StartTime="775.050" EndTime="775.422" Participant="me013">
um.
</Segment>
<Segment StartTime="776.336" EndTime="776.834" Participant="me011">
Even,
</Segment>
<Segment StartTime="776.834" EndTime="778.767" Participant="me011">
I guess with vowels that would be pretty hard,
</Segment>
<Segment StartTime="778.767" EndTime="779.344" Participant="me011">
wouldn't it?
</Segment>
<Segment StartTime="782.512" EndTime="785.043" Participant="me011">
To identify actually, you know, which one it is?
</Segment>
<Segment StartTime="785.522" EndTime="788.899" Participant="fe008">
It would seem to me that the points of articulation would be m- more,
</Segment>
<Segment StartTime="789.965" EndTime="792.979" Participant="fe008">
g- uh, I mean that's - I think about articulatory features, I think about,
</Segment>
<Segment StartTime="792.979" EndTime="794.590" Participant="fe008">
points of articulation, which means,
</Segment>
<Segment StartTime="795.319" EndTime="796.819" Participant="fe008">
uh, rather than vowels.
</Segment>
<Segment StartTime="795.529" EndTime="796.027" Participant="me011">
Yeah.
</Segment>
<Segment StartTime="797.341" EndTime="798.892" Participant="me018">
<Emphasis> Points </Emphasis> of articulation? What do you mean?
</Segment>
<Segment StartTime="799.157" EndTime="803.012" Participant="fe008">
So, is it, uh, bilabial or dental or is it, you know,
</Segment>
<Segment StartTime="803.405" EndTime="804.092" Participant="fe008">
palatal.
</Segment>
<Segment StartTime="803.589" EndTime="804.034" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="804.179" EndTime="806.112" Participant="mn014">
<Uncertain NumSyllables="4"> @@ </Uncertain> <Comment Description="quietly"/>
</Segment>
<Segment StartTime="804.737" EndTime="807.323" Participant="fe008">
Which - which are all like where - where your tongue comes to rest.
</Segment>
<Segment StartTime="805.556" EndTime="806.990" Participant="me011">
Uvular.
</Segment>
<Segment StartTime="806.262" EndTime="807.935" Participant="me018">
Place of ar- place of articulation.
</Segment>
<Segment StartTime="806.510" EndTime="807.383" Participant="me013">
<Emphasis> Place, </Emphasis> place.
</Segment>
<Segment StartTime="807.763" EndTime="809.198" Participant="mn014">
Place. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="807.882" EndTime="811.017" Participant="fe008">
Place. Thank you, what - whatev- whatever I s- said, that's - I really meant place.
</Segment>
<Segment StartTime="807.944" EndTime="808.286" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="808.593" EndTime="808.952" Participant="me018">
O_K.
</Segment>
<Segment StartTime="810.001" EndTime="810.272" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="810.207" EndTime="811.122" Participant="me018">
O_K, I see.
</Segment>
<Segment StartTime="811.017" EndTime="811.886" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="811.578" EndTime="811.877" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="812.052" EndTime="813.497" Participant="me013">
O_K we got our jargon then, O_K.
</Segment>
<Segment StartTime="813.165" EndTime="813.536" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="813.511" EndTime="813.875" Participant="fe008">
Yeah.
</Segment>
<Segment StartTime="814.300" EndTime="814.912" Participant="me013">
Uh.
</Segment>
<Segment StartTime="815.307" EndTime="816.653" Participant="fe016">
Well it's also, there's,
</Segment>
<Segment StartTime="817.321" EndTime="819.093" Participant="fe016">
really a difference between,
</Segment>
<Segment StartTime="819.093" EndTime="821.802" Participant="fe016">
the pronunciation models in the dictionary, and,
</Segment>
<Segment StartTime="821.802" EndTime="823.695" Participant="fe016">
the pronunciations that people produce.
</Segment>
<Segment StartTime="823.695" EndTime="824.273" Participant="fe016">
And, so,
</Segment>
<Segment StartTime="824.552" EndTime="825.030" Participant="fe016">
You get,
</Segment>
<Segment StartTime="825.030" EndTime="827.839" Participant="fe016">
some of that information from Steve's work on the -
</Segment>
<Segment StartTime="827.680" EndTime="827.983" Participant="me013">
Right.
</Segment>
<Segment StartTime="827.839" EndTime="829.379" Participant="fe016">
on the labeling and it really,
</Segment>
<Segment StartTime="827.849" EndTime="828.290" Participant="me011">
Right.
</Segment>
<Segment StartTime="830.529" EndTime="833.673" Participant="fe016">
I actually think that data should be used more. That maybe,
</Segment>
<Segment StartTime="834.370" EndTime="840.284" Participant="fe016">
although I think the meeting context is great, that he has transcriptions that give you the actual phone sequence.
</Segment>
<Segment StartTime="840.881" EndTime="842.132" Participant="fe016">
And you can go from -
</Segment>
<Segment StartTime="842.371" EndTime="846.874" Participant="fe016">
not from that to the articulatory features, but that would be a better starting point for marking,
</Segment>
<Segment StartTime="847.352" EndTime="848.987" Participant="fe016">
the gestural features, then,
</Segment>
<Segment StartTime="849.362" EndTime="851.294" Participant="fe016">
data where you don't have that, because,
</Segment>
<Segment StartTime="852.011" EndTime="853.354" Participant="fe016">
we - you wanna know,
</Segment>
<Segment StartTime="854.111" EndTime="856.953" Participant="fe016">
both about the way that they're producing a certain sound,
</Segment>
<Segment StartTime="857.272" EndTime="858.454" Participant="fe016">
and what kinds of,
</Segment>
<Segment StartTime="859.290" EndTime="859.938" Participant="fe016">
<Uncertain> you know </Uncertain> what kinds of,
</Segment>
<Segment StartTime="861.010" EndTime="861.653" Participant="fe016">
phonemic,
</Segment>
<Segment StartTime="862.191" EndTime="864.117" Participant="fe016">
differences you get between these,
</Segment>
<Segment StartTime="864.277" EndTime="865.233" Participant="fe016">
transcribed,
</Segment>
<Segment StartTime="865.550" EndTime="867.613" Participant="fe016">
sequences and the dictionary ones.
</Segment>
<Segment StartTime="867.597" EndTime="870.166" Participant="me013">
Well you might be right that mi- might be the way at <Emphasis> getting </Emphasis> at,
</Segment>
<Segment StartTime="870.166" EndTime="871.365" Participant="me013">
what I was talking about,
</Segment>
<Segment StartTime="871.365" EndTime="875.240" Participant="me013">
but the particular reason why I was interested in doing that was because I <Emphasis> remember, </Emphasis>
</Segment>
<Segment StartTime="875.656" EndTime="877.440" Participant="me013">
when that happened, and,
</Segment>
<Segment StartTime="877.564" EndTime="881.870" Participant="me013">
John Ohala <Emphasis> was </Emphasis> over here and he was <Emphasis> looking </Emphasis> at the spectrograms of the more difficult ones.
</Segment>
<Segment StartTime="882.789" EndTime="884.723" Participant="me013">
Uh, he didn't know what to say,
</Segment>
<Segment StartTime="885.306" EndTime="885.999" Participant="me013">
about,
</Segment>
<Segment StartTime="886.655" EndTime="888.414" Participant="me013">
what is the sequence of phones there.
</Segment>
<Segment StartTime="888.414" EndTime="889.903" Participant="me013">
They came up with some compromise.
</Segment>
<Segment StartTime="890.232" EndTime="896.349" Participant="me013">
Because that really wasn't what it look like. It didn't look like a sequence of phones it look like this blending thing happening here and here and here.
</Segment>
<Segment StartTime="891.710" EndTime="892.273" Participant="fe016">
Right.
</Segment>
<Segment StartTime="891.990" EndTime="892.473" Participant="me011">
Right.
</Segment>
<Segment StartTime="895.542" EndTime="896.040" Participant="fe016">
Right.
</Segment>
<Segment StartTime="895.662" EndTime="897.671" Participant="me011">
Yeah, so you have this feature here, and,
</Segment>
<Segment StartTime="896.889" EndTime="898.130" Participant="me018">
There was no <Emphasis> name </Emphasis> for that.
</Segment>
<Segment StartTime="897.069" EndTime="897.493" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="897.202" EndTime="897.504" Participant="fe016">
But -
</Segment>
<Segment StartTime="897.671" EndTime="898.655" Participant="me011">
overlap, yeah.
</Segment>
<Segment StartTime="897.883" EndTime="899.656" Participant="fe016">
Right. But it still is - there's a -
</Segment>
<Segment StartTime="898.130" EndTime="898.361" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="899.895" EndTime="901.110" Participant="fe016">
there are two steps. One -
</Segment>
<Segment StartTime="901.110" EndTime="904.159" Participant="fe016">
you know, one is going from a dictionary pronunciation of something,
</Segment>
<Segment StartTime="901.205" EndTime="901.703" Participant="me011">
And -
</Segment>
<Segment StartTime="904.262" EndTime="904.654" Participant="me013">
Right.
</Segment>
<Segment StartTime="904.856" EndTime="905.474" Participant="fe016">
like,
</Segment>
<Segment StartTime="906.868" EndTime="910.561" Participant="fe016">
"gonna see you tomorrow", it could be "going to" or "gonna" or "<Uncertain> gonta </Uncertain> s-" you know.
</Segment>
<Segment StartTime="906.894" EndTime="907.207" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="910.038" EndTime="910.337" Participant="me013">
Right.
</Segment>
<Segment StartTime="910.659" EndTime="912.340" Participant="me011">
Or "<Uncertain> gonta </Uncertain>". <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="910.880" EndTime="911.340" Participant="fe016">
And,
</Segment>
<Segment StartTime="911.340" EndTime="912.274" Participant="fe016">
yeah. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="911.670" EndTime="912.807" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="911.702" EndTime="912.298" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="912.274" EndTime="913.589" Participant="fe016">
"Gonna see you tomorrow", uh,
</Segment>
<Segment StartTime="913.589" EndTime="914.605" Participant="fe016">
"<Uncertain> guh </Uncertain> see you tomorrow".
</Segment>
<Segment StartTime="914.605" EndTime="915.123" Participant="fe016">
And,
</Segment>
<Segment StartTime="915.522" EndTime="916.926" Participant="fe016">
that it would be nice to have these,
</Segment>
<Segment StartTime="917.471" EndTime="923.425" Participant="fe016">
intermediate, or these - some - these reduced pronunciations that those transcribers had marked or to have people mark those as well.
</Segment>
<Segment StartTime="920.915" EndTime="921.371" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="924.082" EndTime="925.234" Participant="fe016">
Because, it's not,
</Segment>
<Segment StartTime="925.592" EndTime="926.150" Participant="fe016">
um,
</Segment>
<Segment StartTime="926.668" EndTime="928.534" Participant="fe016">
that easy to go from the,
</Segment>
<Segment StartTime="928.912" EndTime="929.727" Participant="fe016">
dictionary,
</Segment>
<Segment StartTime="930.088" EndTime="932.669" Participant="fe016">
word pronuncia- the dictionary phone pronunciation,
</Segment>
<Segment StartTime="932.968" EndTime="936.973" Participant="fe016">
to the gestural one without this intermediate or a syllable level kind of,
</Segment>
<Segment StartTime="937.332" EndTime="938.386" Participant="fe016">
representation.
</Segment>
<Segment StartTime="937.585" EndTime="939.963" Participant="me011">
Well I don't think Morgan's suggesting that we <Emphasis> do </Emphasis> that, though.
</Segment>
<Segment StartTime="938.411" EndTime="938.842" Participant="me013">
Do you mean,
</Segment>
<Segment StartTime="940.684" EndTime="941.063" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="940.761" EndTime="943.959" Participant="me013">
Yeah, I mean, I- I- I'm jus- at the moment of course we're just talking about what,
</Segment>
<Segment StartTime="944.144" EndTime="948.085" Participant="me013">
to provide as a tool for people to do research who have different ideas about how to do it.
</Segment>
<Segment StartTime="945.153" EndTime="946.187" Participant="me018">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="948.085" EndTime="948.940" Participant="me013">
So for instance,
</Segment>
<Segment StartTime="949.391" EndTime="952.498" Participant="me013">
you might have someone who just has a wor- has words with states,
</Segment>
<Segment StartTime="952.858" EndTime="954.387" Participant="me013">
and has uh - uh,
</Segment>
<Segment StartTime="954.585" EndTime="956.568" Participant="me013">
comes from articulatory gestures to that.
</Segment>
<Segment StartTime="956.568" EndTime="957.409" Participant="me013">
And someone else,
</Segment>
<Segment StartTime="957.711" EndTime="960.424" Participant="me013">
might actually want some phonetic uh intermediate thing.
</Segment>
<Segment StartTime="960.869" EndTime="963.639" Participant="me013">
So I think it would be - be best to have all of it if we could.
</Segment>
<Segment StartTime="964.267" EndTime="965.138" Participant="me013">
But <Pause/> um,
</Segment>
<Segment StartTime="964.904" EndTime="965.203" Participant="me011">
But -
</Segment>
<Segment StartTime="965.602" EndTime="968.004" Participant="me011">
What I'm imagining is a score-like notation,
</Segment>
<Segment StartTime="965.633" EndTime="965.912" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="968.382" EndTime="970.335" Participant="me011">
where each line is a particular <Emphasis> feature. </Emphasis>
</Segment>
<Segment StartTime="970.653" EndTime="975.913" Participant="me011">
Right, so you would say, you know, it's voiced through here, and so you have label here, and you have nas- nasal here, and,
</Segment>
<Segment StartTime="971.171" EndTime="971.505" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="976.212" EndTime="981.064" Participant="me011">
they - they could be overlapping in all sorts of bizarre ways that don't correspond to the timing on phones.
</Segment>
<Segment StartTime="981.564" EndTime="984.361" Participant="me013">
I mean this is the kind of reason why - I remember when at one of the Switchboard,
</Segment>
<Segment StartTime="984.510" EndTime="988.533" Participant="me013">
workshops, that uh when we <Emphasis> talked </Emphasis> about doing the transcription project,
</Segment>
<Segment StartTime="988.964" EndTime="990.710" Participant="me013">
Dave Talkin said, "can't be done".
</Segment>
<Segment StartTime="990.710" EndTime="991.842" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="990.943" EndTime="991.442" Participant="me011">
Right.
</Segment>
<Segment StartTime="991.842" EndTime="994.597" Participant="me013">
He was - he was, what - what he meant was that this isn't,
</Segment>
<Segment StartTime="994.990" EndTime="997.980" Participant="me013">
you know, a sequence of phones, and when you actually look at Switchboard that's,
</Segment>
<Segment StartTime="998.061" EndTime="999.119" Participant="me013">
not what you see, and, you know.
</Segment>
<Segment StartTime="998.889" EndTime="1002.617" Participant="me011">
And in - in fact the inter-annotator <Emphasis> agreement </Emphasis> was not that good, right?
</Segment>
<Segment StartTime="999.550" EndTime="1000.002" Participant="me013">
And.
</Segment>
<Segment StartTime="1002.936" EndTime="1003.997" Participant="me011">
On the harder ones?
</Segment>
<Segment StartTime="1004.928" EndTime="1006.155" Participant="me013">
It, yeah I mean it was-
</Segment>
<Segment StartTime="1005.147" EndTime="1009.564" Participant="fe016">
It depends how you look at it, and I - I understand what you're saying about this,
</Segment>
<Segment StartTime="1009.660" EndTime="1009.956" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1009.787" EndTime="1014.031" Participant="fe016">
kind of transcription exactly, because I've seen - you know, where does the voicing bar start and so forth.
</Segment>
<Segment StartTime="1012.250" EndTime="1012.666" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1014.390" EndTime="1015.486" Participant="fe016">
All I'm saying is that,
</Segment>
<Segment StartTime="1015.743" EndTime="1018.114" Participant="fe016">
it is useful to have that -
</Segment>
<Segment StartTime="1018.452" EndTime="1022.623" Participant="fe016">
the transcription of what was <Emphasis> really </Emphasis> said, and which syllables were reduced.
</Segment>
<Segment StartTime="1023.042" EndTime="1028.247" Participant="fe016">
Uh, if you're gonna add the features it's also useful to have some level of representation which is,
</Segment>
<Segment StartTime="1029.638" EndTime="1030.790" Participant="fe016">
is a reduced -
</Segment>
<Segment StartTime="1030.790" EndTime="1032.842" Participant="fe016">
it's a pronunciation variant,
</Segment>
<Segment StartTime="1032.842" EndTime="1037.425" Participant="fe016">
that currently the dictionaries don't give you because if you add them to the dictionary and you run recognition, you,
</Segment>
<Segment StartTime="1033.512" EndTime="1034.010" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="1037.488" EndTime="1037.798" Participant="me013">
Right.
</Segment>
<Segment StartTime="1037.684" EndTime="1038.700" Participant="fe016">
you add confusion.
</Segment>
<Segment StartTime="1038.700" EndTime="1040.453" Participant="fe016">
So people purposely don't add them.
</Segment>
<Segment StartTime="1038.897" EndTime="1039.173" Participant="me013">
Right.
</Segment>
<Segment StartTime="1041.549" EndTime="1043.922" Participant="fe016">
So it's useful to know which variant was -
</Segment>
<Segment StartTime="1043.922" EndTime="1046.260" Participant="fe016">
was produced, at least at the phone level.
</Segment>
<Segment StartTime="1045.174" EndTime="1047.243" Participant="me018">
So it would be - it would be great if we had,
</Segment>
<Segment StartTime="1047.602" EndTime="1049.353" Participant="me018">
either these kind of,
</Segment>
<Segment StartTime="1049.700" EndTime="1050.907" Participant="me018">
labelings on,
</Segment>
<Segment StartTime="1051.341" EndTime="1053.841" Participant="me018">
the same portion of Switchboard that Steve marked, or,
</Segment>
<Segment StartTime="1054.414" EndTime="1056.765" Participant="me018">
Steve's type markings on this data,
</Segment>
<Segment StartTime="1055.007" EndTime="1055.650" Participant="fe016">
Right.
</Segment>
<Segment StartTime="1055.650" EndTime="1056.327" Participant="fe016">
That's all, <Uncertain> I mean </Uncertain>.
</Segment>
<Segment StartTime="1056.506" EndTime="1057.263" Participant="fe016">
Exactly.
</Segment>
<Segment StartTime="1056.765" EndTime="1057.502" Participant="me018">
with these.
</Segment>
<Segment StartTime="1057.110" EndTime="1057.402" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1057.642" EndTime="1058.279" Participant="fe016">
Exactly.
</Segment>
<Segment StartTime="1058.225" EndTime="1059.771" Participant="me013">
Yeah, no I - I don't disagree with that.
</Segment>
<Segment StartTime="1058.817" EndTime="1061.652" Participant="fe016">
And Steve's type is fairly - it's not that slow,
</Segment>
<Segment StartTime="1061.652" EndTime="1062.064" Participant="fe016">
uh,
</Segment>
<Segment StartTime="1062.502" EndTime="1063.843" Participant="fe016">
uh, I dunno exactly what the,
</Segment>
<Segment StartTime="1064.342" EndTime="1065.444" Participant="fe016">
timing was, but.
</Segment>
<Segment StartTime="1064.431" EndTime="1066.986" Participant="me013">
Yeah u- I don't disagree with it the on- the only thing is that,
</Segment>
<Segment StartTime="1067.127" EndTime="1069.867" Participant="me013">
What you actually will end - en- end up with is something,
</Segment>
<Segment StartTime="1069.867" EndTime="1071.544" Participant="me013">
i- it's all compromised, right, so,
</Segment>
<Segment StartTime="1071.935" EndTime="1074.288" Participant="me013">
the string that you end up with isn't,
</Segment>
<Segment StartTime="1074.565" EndTime="1075.223" Participant="me013">
actually,
</Segment>
<Segment StartTime="1075.621" EndTime="1076.411" Participant="me013">
what happened.
</Segment>
<Segment StartTime="1076.499" EndTime="1081.429" Participant="me013">
But it's - it's the best <Emphasis> compromise </Emphasis> that a group of people scratching their heads could come up with to describe what happened.
</Segment>
<Segment StartTime="1081.469" EndTime="1081.947" Participant="fe016">
Mm-hmm.
</Segment>
<Segment StartTime="1081.700" EndTime="1083.251" Participant="me018">
And it's more accurate than,
</Segment>
<Segment StartTime="1082.440" EndTime="1082.736" Participant="me013">
But.
</Segment>
<Segment StartTime="1083.140" EndTime="1085.769" Participant="me013">
And it's more accurate than the - than the dictionary or,
</Segment>
<Segment StartTime="1083.251" EndTime="1084.288" Participant="me018">
<Emphasis> phone </Emphasis> labels.
</Segment>
<Segment StartTime="1083.481" EndTime="1084.260" Participant="me011">
The word. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1085.185" EndTime="1085.523" Participant="me018">
Yeah.
</Segment>
<Segment StartTime="1086.146" EndTime="1093.188" Participant="me013">
if you've got a pronunciation uh lexicon that has three or four, this might be have been the fifth one that you <Uncertain> tr- </Uncertain> that you pruned or whatever, so sure.
</Segment>
<Segment StartTime="1091.042" EndTime="1094.057" Participant="me018">
So it's like a continuum. It's - you're going all the way down, yeah.
</Segment>
<Segment StartTime="1091.108" EndTime="1091.648" Participant="fe016">
Right.
</Segment>
<Segment StartTime="1091.927" EndTime="1092.365" Participant="fe016">
Right.
</Segment>
<Segment StartTime="1092.724" EndTime="1093.092" Participant="fe016">
Right.
</Segment>
<Segment StartTime="1093.411" EndTime="1096.647" Participant="fe016">
That's what I meant is - an- and in some places it would fill in,
</Segment>
<Segment StartTime="1093.914" EndTime="1094.183" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1094.883" EndTime="1095.214" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1096.698" EndTime="1097.154" Participant="me011">
Well -
</Segment>
<Segment StartTime="1096.899" EndTime="1099.585" Participant="fe016">
So - the kinds of gestural features are not everywhere.
</Segment>
<Segment StartTime="1099.805" EndTime="1103.693" Participant="fe016">
So there are some things that you don't have access to either from your ear <Emphasis> or </Emphasis> the spectrogram,
</Segment>
<Segment StartTime="1099.808" EndTime="1100.153" Participant="me018">
Right.
</Segment>
<Segment StartTime="1103.753" EndTime="1104.251" Participant="me018">
Mm-hmm.
</Segment>
<Segment StartTime="1103.952" EndTime="1106.264" Participant="fe016">
but you know what phone it was and that's about all you can -
</Segment>
<Segment StartTime="1106.622" EndTime="1108.670" Participant="fe016">
all you can say. And then there are <Emphasis> other </Emphasis> cases where,
</Segment>
<Segment StartTime="1106.741" EndTime="1107.100" Participant="me018">
Right.
</Segment>
<Segment StartTime="1108.555" EndTime="1109.616" Participant="me018">
It's basically just having,
</Segment>
<Segment StartTime="1109.048" EndTime="1110.323" Participant="fe016">
nasality, voicing -
</Segment>
<Segment StartTime="1109.616" EndTime="1111.091" Participant="me018">
multiple levels of -
</Segment>
<Segment StartTime="1110.802" EndTime="1111.260" Participant="fe016">
<NonVocalSound Description="mike noise"/> Right.
</Segment>
<Segment StartTime="1111.091" EndTime="1111.549" Participant="me018">
of,
</Segment>
<Segment StartTime="1111.260" EndTime="1112.140" Participant="fe016">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1111.549" EndTime="1112.705" Participant="me018">
information and marking,
</Segment>
<Segment StartTime="1112.140" EndTime="1112.593" Participant="fe016">
Right. <NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1112.736" EndTime="1113.329" Participant="me013">
Yeah. <VocalSound Description="glottal sound"/>
</Segment>
<Segment StartTime="1112.983" EndTime="1113.680" Participant="me018">
on the signal.
</Segment>
<Segment StartTime="1113.010" EndTime="1115.913" Participant="me011">
Well the other difference is that the - the <Emphasis> features, </Emphasis>
</Segment>
<Segment StartTime="1113.051" EndTime="1113.310" Participant="fe016">
<NonVocalSound Description="mike noise"/> Right.
</Segment>
<Segment StartTime="1113.160" EndTime="1113.651" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1113.310" EndTime="1123.571" Participant="fe016">
<Pause/> <NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1116.172" EndTime="1119.634" Participant="me011">
are not synchronous, right. They overlap each other in weird ways.
</Segment>
<Segment StartTime="1118.641" EndTime="1119.080" Participant="me018">
Mm-hmm.
</Segment>
<Segment StartTime="1119.595" EndTime="1119.994" Participant="me018">
Mm-hmm.
</Segment>
<Segment StartTime="1119.998" EndTime="1122.068" Participant="me011">
So it's not a strictly one-dimensional signal.
</Segment>
<Segment StartTime="1122.424" EndTime="1122.750" Participant="me013">
Right.
</Segment>
<Segment StartTime="1122.897" EndTime="1125.069" Participant="me011">
So I think that's sorta qualitatively different.
</Segment>
<Segment StartTime="1123.571" EndTime="1125.564" Participant="fe016">
Right. You can add the features in,
</Segment>
<Segment StartTime="1125.982" EndTime="1131.044" Participant="fe016">
uh, but it'll be underspecified. Th- <Pronounce Pronunciation="&quot;thell&quot;"> there'll </Pronounce> be no way for you to actually mark what was said completely by features.
</Segment>
<Segment StartTime="1126.481" EndTime="1126.867" Participant="fe008">
Hmm.
</Segment>
<Segment StartTime="1132.023" EndTime="1135.350" Participant="me011">
Well not with our current system but you could imagine designing a system,
</Segment>
<Segment StartTime="1132.139" EndTime="1132.992" Participant="fe016">
And i- if you're -
</Segment>
<Segment StartTime="1135.659" EndTime="1135.918" Participant="fe016">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1136.007" EndTime="1138.532" Participant="me011">
that the states were features, rather than phones.
</Segment>
<Segment StartTime="1140.142" EndTime="1140.682" Participant="fe016">
Well,
</Segment>
<Segment StartTime="1140.593" EndTime="1141.028" Participant="fe008">
That's -
</Segment>
<Segment StartTime="1140.901" EndTime="1142.794" Participant="fe016">
we - we've probably have a <VocalSound Description="laugh"/> separate,
</Segment>
<Segment StartTime="1142.794" EndTime="1143.232" Participant="fe016">
um,
</Segment>
<Segment StartTime="1142.879" EndTime="1143.496" Participant="me011">
Yeah. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1143.232" EndTime="1145.404" Participant="fe016">
discussion of, uh - of whether you can do that.
</Segment>
<Segment StartTime="1143.594" EndTime="1145.192" Participant="fe008">
Well, <Pause/> isn't that - I thought that was,
</Segment>
<Segment StartTime="1145.487" EndTime="1147.550" Participant="fe008">
well but that - wasn't that kinda the direction? I thought
</Segment>
<Segment StartTime="1147.516" EndTime="1148.379" Participant="me013">
Yeah, so I mean, what,
</Segment>
<Segment StartTime="1148.379" EndTime="1150.063" Participant="me013">
what - where this is, I mean,
</Segment>
<Segment StartTime="1150.063" EndTime="1155.710" Participant="me013">
I- I want- would like to have something that's useful to people other than those who are doing the specific kind of research I have in mind,
</Segment>
<Segment StartTime="1155.710" EndTime="1157.075" Participant="me013">
so it should be something broader. But,
</Segment>
<Segment StartTime="1157.566" EndTime="1160.453" Participant="me013">
The - <Emphasis> but </Emphasis> uh where I'm coming from is, uh,
</Segment>
<Segment StartTime="1160.890" EndTime="1164.911" Participant="me013">
we're coming off of stuff that <Uncertain> Larry Saul </Uncertain> did with - with, um,
</Segment>
<Segment StartTime="1165.755" EndTime="1168.826" Participant="me013">
uh, <Uncertain> John Dalan </Uncertain> and <Uncertain> Muzim Rahim </Uncertain> in which, uh, they,
</Segment>
<Segment StartTime="1169.320" EndTime="1170.921" Participant="me013">
uh, have, um,
</Segment>
<Segment StartTime="1171.781" EndTime="1177.836" Participant="me013">
a m- a multi-band system that is, uh, trained through a combination of gradient learning an- and E_M,
</Segment>
<Segment StartTime="1178.281" EndTime="1179.393" Participant="me013">
to <Pause/> um,
</Segment>
<Segment StartTime="1179.916" EndTime="1181.289" Participant="me013">
estimate, uh,
</Segment>
<Segment StartTime="1181.289" EndTime="1181.768" Participant="me013">
<VocalSound Description="glottal sound"/>
</Segment>
<Segment StartTime="1181.768" EndTime="1182.228" Participant="me013">
the,
</Segment>
<Segment StartTime="1182.907" EndTime="1185.500" Participant="me013">
uh, value for m- for - for a particular feature.
</Segment>
<Segment StartTime="1185.500" EndTime="1185.862" Participant="me013">
O_K.
</Segment>
<Segment StartTime="1186.248" EndTime="1188.066" Participant="me013">
And this is part of a larger,
</Segment>
<Segment StartTime="1188.066" EndTime="1192.358" Participant="me013">
image that <Uncertain> John Dalan </Uncertain> has about how the human brain does it in which he's sort of imagining that,
</Segment>
<Segment StartTime="1192.769" EndTime="1196.007" Participant="me013">
individual frequency channels are coming up with their own estimate,
</Segment>
<Segment StartTime="1196.581" EndTime="1197.663" Participant="me013">
of - of these,
</Segment>
<Segment StartTime="1197.818" EndTime="1200.830" Participant="me013">
these kinds of - something like this. Might not be, you know, exact features that,
</Segment>
<Segment StartTime="1200.830" EndTime="1202.804" Participant="me013">
Jakobson thought of or something. But I mean you know some,
</Segment>
<Segment StartTime="1202.897" EndTime="1203.468" Participant="me013">
something like that.
</Segment>
<Segment StartTime="1203.468" EndTime="1205.213" Participant="me013">
Some kind of low-level features, which are <Emphasis> not, </Emphasis>
</Segment>
<Segment StartTime="1205.594" EndTime="1207.391" Participant="me013">
fully, you know, phone classification.
</Segment>
<Segment StartTime="1207.928" EndTime="1210.792" Participant="me013">
And the - the - th- this particular image,
</Segment>
<Segment StartTime="1211.086" EndTime="1212.376" Participant="me013">
of how <Uncertain> thi- </Uncertain> how it's done,
</Segment>
<Segment StartTime="1212.376" EndTime="1213.034" Participant="me013">
is that,
</Segment>
<Segment StartTime="1213.164" EndTime="1215.935" Participant="me013">
then given all of these estimates at that level,
</Segment>
<Segment StartTime="1216.311" EndTime="1217.425" Participant="me013">
there's a level above it,
</Segment>
<Segment StartTime="1217.524" EndTime="1219.287" Participant="me013">
then which is - is making,
</Segment>
<Segment StartTime="1219.602" EndTime="1223.243" Participant="me013">
some kind of sound unit classification such as, you know, phone and - and, you know.
</Segment>
<Segment StartTime="1223.769" EndTime="1224.622" Participant="me013">
You could argue what,
</Segment>
<Segment StartTime="1224.827" EndTime="1228.568" Participant="me013">
what a sound unit should be, and - and so forth. But that - that's sort of what I was imagining doing,
</Segment>
<Segment StartTime="1229.067" EndTime="1230.228" Participant="me013">
um, and -
</Segment>
<Segment StartTime="1230.548" EndTime="1235.365" Participant="me013">
but it's still open within that whether you would have an intermediate level in which it was actually phones, or <Emphasis> not. </Emphasis>
</Segment>
<Segment StartTime="1235.365" EndTime="1236.546" Participant="me013">
You wouldn't necessarily have to.
</Segment>
<Segment StartTime="1237.390" EndTime="1238.119" Participant="me013">
Um, but,
</Segment>
<Segment StartTime="1238.457" EndTime="1239.282" Participant="me013">
Again, I wouldn't wanna,
</Segment>
<Segment StartTime="1240.025" EndTime="1245.796" Participant="me013">
wouldn't want what we - we produced to be so, know, local in perspective that it - it was matched,
</Segment>
<Segment StartTime="1246.170" EndTime="1247.790" Participant="me013">
what we were thinking of doing one week,
</Segment>
<Segment StartTime="1247.790" EndTime="1248.359" Participant="me013">
And - and,
</Segment>
<Segment StartTime="1248.359" EndTime="1250.568" Participant="me013">
and, you know, what you're saying is absolutely right. That,
</Segment>
<Segment StartTime="1250.663" EndTime="1251.383" Participant="me013">
that if we,
</Segment>
<Segment StartTime="1251.846" EndTime="1253.171" Participant="me013">
can we should put in,
</Segment>
<Segment StartTime="1253.171" EndTime="1254.527" Participant="me013">
uh, another level of,
</Segment>
<Segment StartTime="1254.527" EndTime="1257.289" Participant="me013">
of description there if we're gonna get into some of this low-level stuff.
</Segment>
<Segment StartTime="1257.315" EndTime="1259.574" Participant="me018">
Well, you know, um - I mean if we're talking about,
</Segment>
<Segment StartTime="1259.574" EndTime="1260.767" Participant="me018">
having the,
</Segment>
<Segment StartTime="1261.125" EndTime="1264.363" Participant="me018">
annotators annotate these kinds of features, it seems like,
</Segment>
<Segment StartTime="1265.419" EndTime="1265.977" Participant="me018">
You know, you -
</Segment>
<Segment StartTime="1266.316" EndTime="1267.534" Participant="me018">
The - the question is,
</Segment>
<Segment StartTime="1267.534" EndTime="1268.811" Participant="me018">
do they do that on,
</Segment>
<Segment StartTime="1268.811" EndTime="1269.608" Participant="me018">
meeting data?
</Segment>
<Segment StartTime="1269.608" EndTime="1270.783" Participant="me018">
Or do they do that on,
</Segment>
<Segment StartTime="1271.142" EndTime="1271.857" Participant="me018">
Switchboard?
</Segment>
<Segment StartTime="1271.608" EndTime="1274.514" Participant="me011">
That's what I was saying, maybe meeting data <Emphasis> isn't </Emphasis> the right corpus.
</Segment>
<Segment StartTime="1272.219" EndTime="1272.808" Participant="fe008">
<VocalSound Description="inbreath"/> W-
</Segment>
<Segment StartTime="1274.931" EndTime="1278.520" Participant="fe008">
Well it seems like you could do both. I mean, I was thinking that it would be interesting,
</Segment>
<Segment StartTime="1277.749" EndTime="1278.435" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="1278.520" EndTime="1279.782" Participant="fe008">
to do it with respect to,
</Segment>
<Segment StartTime="1279.782" EndTime="1283.595" Participant="fe008">
parts of Switchboard anyway, in terms of, uh - partly to see,
</Segment>
<Segment StartTime="1283.324" EndTime="1283.833" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="1283.595" EndTime="1284.759" Participant="fe008">
if you could,
</Segment>
<Segment StartTime="1284.759" EndTime="1292.260" Participant="fe008">
generate first <Emphasis> guesses </Emphasis> at what the articulatory feature would be, <Emphasis> based </Emphasis> on the phone representation at that lower level. It might be a time gain.
</Segment>
<Segment StartTime="1292.260" EndTime="1294.740" Participant="fe008">
But also in terms of comparability of, um,
</Segment>
<Segment StartTime="1292.475" EndTime="1292.937" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="1295.416" EndTime="1298.734" Participant="me018">
Well cuz the- yeah, and then also, if you did it on Switchboard, you would have,
</Segment>
<Segment StartTime="1298.852" EndTime="1299.594" Participant="fe008">
what you gain <Comment Description="whispered"/>
</Segment>
<Segment StartTime="1299.093" EndTime="1301.793" Participant="me018">
the full continuum of transcriptions. You'd have it,
</Segment>
<Segment StartTime="1301.094" EndTime="1301.444" Participant="fe008">
Yep.
</Segment>
<Segment StartTime="1301.793" EndTime="1304.762" Participant="me018">
from the lowest level, the ac- acoustic features, then you'd have the,
</Segment>
<Segment StartTime="1305.280" EndTime="1307.491" Participant="me018">
you know, the phonetic level that Steve did, and,
</Segment>
<Segment StartTime="1305.328" EndTime="1305.622" Participant="fe008">
Mm-hmm.
</Segment>
<Segment StartTime="1305.880" EndTime="1309.522" Participant="fe016">
Yeah that - that's all I was thinking about. it <Emphasis> is </Emphasis> telephone band, so,
</Segment>
<Segment StartTime="1307.491" EndTime="1308.023" Participant="me018">
yeah.
</Segment>
<Segment StartTime="1307.557" EndTime="1308.500" Participant="fe008">
And you could tell that -
</Segment>
<Segment StartTime="1308.439" EndTime="1309.763" Participant="me018">
It'd be a <Emphasis> complete, </Emphasis>
</Segment>
<Segment StartTime="1309.604" EndTime="1311.749" Participant="fe008">
And you get the relative gain <Uncertain> up ahead </Uncertain>.
</Segment>
<Segment StartTime="1309.721" EndTime="1311.136" Participant="fe016">
the bandwidth might be -
</Segment>
<Segment StartTime="1309.763" EndTime="1310.460" Participant="me018">
set then.
</Segment>
<Segment StartTime="1311.288" EndTime="1312.536" Participant="me013">
It's so it's a little different.
</Segment>
<Segment StartTime="1312.730" EndTime="1313.413" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="1312.790" EndTime="1315.056" Participant="me013">
So I mean i- we'll see wha- how much we can,
</Segment>
<Segment StartTime="1315.100" EndTime="1315.520" Participant="fe008">
Mm-hmm.
</Segment>
<Segment StartTime="1315.194" EndTime="1315.633" Participant="me013">
uh,
</Segment>
<Segment StartTime="1315.834" EndTime="1319.114" Participant="me013">
get the people to do, and how much money we'll have and all this sort of thing, but,
</Segment>
<Segment StartTime="1317.273" EndTime="1317.539" Participant="fe008">
Mm-hmm.
</Segment>
<Segment StartTime="1318.004" EndTime="1320.878" Participant="me018">
But it - it might be good to do what Jane was saying uh, you know,
</Segment>
<Segment StartTime="1318.256" EndTime="1318.631" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1319.453" EndTime="1320.315" Participant="me013">
Might be do both.
</Segment>
<Segment StartTime="1320.878" EndTime="1322.043" Participant="me018">
<Emphasis> seed </Emphasis> it, with,
</Segment>
<Segment StartTime="1322.043" EndTime="1324.763" Participant="me018">
guesses about what we think the features are, based on,
</Segment>
<Segment StartTime="1325.310" EndTime="1327.584" Participant="me018">
you know, the phone or Steve's transcriptions or something.
</Segment>
<Segment StartTime="1327.843" EndTime="1328.958" Participant="me018">
to make it quicker.
</Segment>
<Segment StartTime="1328.022" EndTime="1331.925" Participant="me011">
Alright, so based on the phone transcripts they would all be synchronous, but then you could imagine,
</Segment>
<Segment StartTime="1331.773" EndTime="1333.713" Participant="me018">
Adjusting? Yeah, exactly.
</Segment>
<Segment StartTime="1331.925" EndTime="1333.338" Participant="me011">
nudging them here and there.
</Segment>
<Segment StartTime="1334.909" EndTime="1336.724" Participant="me018">
Scoot the voicing over a little, because -
</Segment>
<Segment StartTime="1335.061" EndTime="1335.398" Participant="fe008">
Yeah.
</Segment>
<Segment StartTime="1336.227" EndTime="1336.685" Participant="me011">
Right.
</Segment>
<Segment StartTime="1336.808" EndTime="1340.453" Participant="me013">
Well I think what - I mean I'm - I'm a l- little behind in what they're doing,
</Segment>
<Segment StartTime="1340.768" EndTime="1344.160" Participant="me013">
now, and, uh, the stuff they're doing on Switchboard now. But I think that,
</Segment>
<Segment StartTime="1344.160" EndTime="1345.627" Participant="me013">
Steve and the gang are doing,
</Segment>
<Segment StartTime="1346.205" EndTime="1349.858" Participant="me013">
<Emphasis> something </Emphasis> with an automatic system first and then doing some adjustment.
</Segment>
<Segment StartTime="1349.858" EndTime="1351.122" Participant="me013">
As I re- as I recall.
</Segment>
<Segment StartTime="1351.500" EndTime="1357.878" Participant="me013">
So I mean that's probably the right way to go anyway, is to - is to start off with an automatic system with a pretty rich pronunciation dictionary that,
</Segment>
<Segment StartTime="1358.425" EndTime="1358.919" Participant="me013">
that,
</Segment>
<Segment StartTime="1358.919" EndTime="1359.443" Participant="me013">
um,
</Segment>
<Segment StartTime="1359.443" EndTime="1359.743" Participant="me013">
you know,
</Segment>
<Segment StartTime="1360.267" EndTime="1360.999" Participant="me013">
tries,
</Segment>
<Segment StartTime="1360.999" EndTime="1362.124" Participant="me013">
to label it all.
</Segment>
<Segment StartTime="1362.124" EndTime="1362.826" Participant="me013">
And then,
</Segment>
<Segment StartTime="1363.126" EndTime="1364.390" Participant="me013">
people go through and fix it.
</Segment>
<Segment StartTime="1365.498" EndTime="1370.768" Participant="fe008">
So in - in our case you'd think about us s- starting with maybe the regular dictionary entry, and then?
</Segment>
<Segment StartTime="1370.730" EndTime="1371.132" Participant="me013">
Well,
</Segment>
<Segment StartTime="1370.768" EndTime="1371.848" Participant="fe008">
Or <Pause/> would we -
</Segment>
<Segment StartTime="1371.132" EndTime="1374.160" Participant="me013">
regular dictionary, I mean, this is a pretty rich dictionary. It's got,
</Segment>
<Segment StartTime="1374.160" EndTime="1375.901" Participant="me013">
got a fair number of pronunciations in it
</Segment>
<Segment StartTime="1375.393" EndTime="1377.513" Participant="me018">
Or you could start from the - if we were gonna,
</Segment>
<Segment StartTime="1375.730" EndTime="1376.201" Participant="fe008">
But -
</Segment>
<Segment StartTime="1377.969" EndTime="1378.918" Participant="me018">
do the same set,
</Segment>
<Segment StartTime="1379.216" EndTime="1380.910" Participant="me018">
of sentences that Steve had,
</Segment>
<Segment StartTime="1381.408" EndTime="1383.699" Participant="me018">
done, we could start with those transcriptions.
</Segment>
<Segment StartTime="1383.196" EndTime="1383.532" Participant="fe008">
Mm-hmm.
</Segment>
<Segment StartTime="1383.820" EndTime="1384.158" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1383.860" EndTime="1387.283" Participant="fe016">
That's actually what I was thinking, is <Uncertain> tha- </Uncertain> - the problem is when you <Emphasis> run, </Emphasis>
</Segment>
<Segment StartTime="1384.219" EndTime="1384.836" Participant="fe008">
So I was thinking -
</Segment>
<Segment StartTime="1385.121" EndTime="1385.655" Participant="me011">
Right.
</Segment>
<Segment StartTime="1385.679" EndTime="1386.063" Participant="me018">
Yeah.
</Segment>
<Segment StartTime="1388.299" EndTime="1390.424" Participant="fe016">
uh, if you run a regular dictionary,
</Segment>
<Segment StartTime="1390.762" EndTime="1391.314" Participant="fe016">
um,
</Segment>
<Segment StartTime="1391.513" EndTime="1393.127" Participant="fe016">
even if you have variants,
</Segment>
<Segment StartTime="1393.266" EndTime="1394.004" Participant="fe016">
in there,
</Segment>
<Segment StartTime="1394.072" EndTime="1394.469" Participant="me018">
Yeah.
</Segment>
<Segment StartTime="1394.382" EndTime="1395.937" Participant="fe016">
which most people don't, you don't always get,
</Segment>
<Segment StartTime="1396.076" EndTime="1396.514" Participant="fe016">
out,
</Segment>
<Segment StartTime="1396.514" EndTime="1400.101" Participant="fe016">
the actual pronunciations, so that's why the human transcriber's giving you the -
</Segment>
<Segment StartTime="1399.207" EndTime="1399.431" Participant="fe008">
Yeah.
</Segment>
<Segment StartTime="1400.539" EndTime="1401.957" Participant="fe016">
that pronunciation, and so y-
</Segment>
<Segment StartTime="1400.724" EndTime="1403.368" Participant="me013">
Actually maybe they're using phone recognizers.
</Segment>
<Segment StartTime="1401.224" EndTime="1401.545" Participant="fe008">
Oh.
</Segment>
<Segment StartTime="1402.950" EndTime="1409.264" Participant="fe016">
they - they - I thought that they were - we <Emphasis> should </Emphasis> catch up on what Steve is, uh - I think that would be a good i- good idea.
</Segment>
<Segment StartTime="1403.348" EndTime="1403.923" Participant="me011">
They are.
</Segment>
<Segment StartTime="1403.368" EndTime="1404.144" Participant="me013">
Is that what they're doing?
</Segment>
<Segment StartTime="1404.660" EndTime="1405.138" Participant="me013">
Oh, O_K.
</Segment>
<Segment StartTime="1406.899" EndTime="1407.198" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1408.579" EndTime="1411.129" Participant="me013">
Yeah, so I think that i- i- we also don't have, I mean,
</Segment>
<Segment StartTime="1411.129" EndTime="1413.828" Participant="me013">
we've got a good start on it, but we don't have a really good,
</Segment>
<Segment StartTime="1414.306" EndTime="1414.824" Participant="me013">
meeting,
</Segment>
<Segment StartTime="1415.359" EndTime="1418.141" Participant="me013">
recorder or recognizer or transcriber or anything yet, so.
</Segment>
<Segment StartTime="1418.139" EndTime="1418.735" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="1418.536" EndTime="1420.649" Participant="me013">
So, I mean another way to look at this is to,
</Segment>
<Segment StartTime="1420.649" EndTime="1421.401" Participant="me013">
is to, uh,
</Segment>
<Segment StartTime="1421.530" EndTime="1423.817" Participant="me013">
do some stuff on Switchboard which has all this other,
</Segment>
<Segment StartTime="1424.211" EndTime="1424.828" Participant="me013">
stuff to it.
</Segment>
<Segment StartTime="1424.828" EndTime="1426.114" Participant="me013">
And then, um,
</Segment>
<Segment StartTime="1426.604" EndTime="1427.272" Participant="me013">
As we get,
</Segment>
<Segment StartTime="1427.539" EndTime="1431.091" Participant="me013">
further down the road and we can do more things ahead of time, we can,
</Segment>
<Segment StartTime="1428.340" EndTime="1428.778" Participant="me018">
Mm-hmm.
</Segment>
<Segment StartTime="1430.845" EndTime="1431.084" Participant="fe008">
O_K. <Comment Description="whispered"/>
</Segment>
<Segment StartTime="1431.091" EndTime="1432.940" Participant="me013">
do some of the same things to the meeting data.
</Segment>
<Segment StartTime="1432.903" EndTime="1433.323" Participant="me018">
Yeah.
</Segment>
<Segment StartTime="1434.056" EndTime="1437.476" Participant="fe008">
And I'm - and these people might - they - they are, s-
</Segment>
<Segment StartTime="1434.105" EndTime="1434.411" Participant="me013">
Yeah
</Segment>
<Segment StartTime="1437.756" EndTime="1442.666" Participant="fe008">
most of them <Emphasis> are </Emphasis> trained with I_P_A. They'd be able to do phonetic-level coding, or articulatory.
</Segment>
<Segment StartTime="1441.653" EndTime="1443.653" Participant="me018">
Are they busy for the next couple years, or - ? <Comment Description="leading question"/>
</Segment>
<Segment StartTime="1443.341" EndTime="1445.524" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1443.377" EndTime="1444.294" Participant="fe016">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="1443.773" EndTime="1445.442" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1443.938" EndTime="1444.813" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1444.030" EndTime="1445.915" Participant="me018">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1445.442" EndTime="1446.480" Participant="fe008">
Well, you know, I mean they,
</Segment>
<Segment StartTime="1447.027" EndTime="1450.013" Participant="fe008">
they - they're interested in continuing working with us, so - I mean - I,
</Segment>
<Segment StartTime="1450.013" EndTime="1453.134" Participant="fe008">
and this would be up their alley, so, we could - when the - when you d-
</Segment>
<Segment StartTime="1453.849" EndTime="1454.507" Participant="fe008">
meet with,
</Segment>
<Segment StartTime="1454.507" EndTime="1455.824" Participant="fe008">
with John Ohala and find,
</Segment>
<Segment StartTime="1455.824" EndTime="1456.525" Participant="fe008">
<Uncertain> you know </Uncertain> what
</Segment>
<Segment StartTime="1456.834" EndTime="1458.516" Participant="fe008">
taxonomy you want to apply, then,
</Segment>
<Segment StartTime="1458.516" EndTime="1459.133" Participant="fe008">
they'd be,
</Segment>
<Segment StartTime="1459.468" EndTime="1459.876" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1459.596" EndTime="1460.689" Participant="fe008">
good to train onto it.
</Segment>
<Segment StartTime="1460.837" EndTime="1461.130" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1461.270" EndTime="1461.932" Participant="me013">
Anyway, this is,
</Segment>
<Segment StartTime="1462.095" EndTime="1462.426" Participant="fe008">
Yeah.
</Segment>
<Segment StartTime="1462.365" EndTime="1465.115" Participant="me013">
not an urgent thing at all, just it came up.
</Segment>
<Segment StartTime="1464.095" EndTime="1466.387" Participant="me018">
It'd be very interesting though, to have <Pause/> that data.
</Segment>
<Segment StartTime="1465.958" EndTime="1466.408" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="1466.010" EndTime="1467.835" Participant="me011">
I wonder, how would you do a forced alignment?
</Segment>
<Segment StartTime="1466.390" EndTime="1467.221" Participant="fe008">
I think so, too.
</Segment>
<Segment StartTime="1466.567" EndTime="1466.945" Participant="fe016">
Might -
</Segment>
<Segment StartTime="1467.125" EndTime="1467.444" Participant="fe016">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1469.560" EndTime="1470.963" Participant="fe008">
<VocalSound Description="laugh"/> Interesting idea.
</Segment>
<Segment StartTime="1470.858" EndTime="1473.320" Participant="me011">
To - to - I mean, you'd wanna iterate, somehow.
</Segment>
<Segment StartTime="1471.132" EndTime="1471.531" Participant="fe016">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1474.257" EndTime="1475.993" Participant="me011">
Yeah. It's interesting thing to think about.
</Segment>
<Segment StartTime="1474.970" EndTime="1476.052" Participant="fe016">
It might be -
</Segment>
<Segment StartTime="1475.245" EndTime="1475.584" Participant="me018">
Hmm.
</Segment>
<Segment StartTime="1476.570" EndTime="1477.506" Participant="fe016">
I was thinking it might be n-
</Segment>
<Segment StartTime="1476.727" EndTime="1478.600" Participant="me011">
I mean you'd - you'd want models for spreading.
</Segment>
<Segment StartTime="1480.126" EndTime="1481.499" Participant="me018">
Of the f- acoustic features?
</Segment>
<Segment StartTime="1480.903" EndTime="1481.468" Participant="me011">
Yeah.
</Segment>
<Segment StartTime="1481.499" EndTime="1481.950" Participant="me018">
Mm-hmm.
</Segment>
<Segment StartTime="1481.587" EndTime="1482.762" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1481.801" EndTime="1482.190" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="1483.710" EndTime="1483.928" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1483.994" EndTime="1484.268" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="1484.762" EndTime="1486.496" Participant="fe016">
Well it might be neat to do some,
</Segment>
<Segment StartTime="1486.675" EndTime="1487.333" Participant="fe016">
phonetic,
</Segment>
<Segment StartTime="1487.990" EndTime="1489.584" Participant="fe016">
features on these,
</Segment>
<Segment StartTime="1490.062" EndTime="1496.143" Participant="fe016">
<Pronounce Pronunciation="&quot;non..*word&quot;"> nonword </Pronounce> words. <Comment Description="rising"/> Are - are these kinds of words that people never - the "huh"s and the "hmm"s and the "huh-" <VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="1496.058" EndTime="1499.508" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1496.143" EndTime="1496.739" Participant="fe016">
and the uh - <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1496.290" EndTime="1496.973" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1496.739" EndTime="1501.302" Participant="fe016">
These k- No, I'm <Emphasis> serious. </Emphasis> There are all these kinds of <Pause/> functional,
</Segment>
<Segment StartTime="1497.826" EndTime="1498.309" Participant="me018">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1501.302" EndTime="1503.372" Participant="fe016">
uh, elements. I don't know what you call <Pause/> them.
</Segment>
<Segment StartTime="1503.372" EndTime="1507.520" Participant="fe016">
But not just fill pauses but all kinds of ways of <Pause/> interrupting <Comment Description="while laughing"/> and so forth.
</Segment>
<Segment StartTime="1506.796" EndTime="1507.443" Participant="me011">
Uh-huh.
</Segment>
<Segment StartTime="1507.520" EndTime="1510.330" Participant="fe016">
And some of them are, <VocalSound Description="laugh"/> yeah, "uh-huh"s, and "hmm"s, and,
</Segment>
<Segment StartTime="1507.844" EndTime="1509.111" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1507.880" EndTime="1508.937" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1510.330" EndTime="1510.708" Participant="fe016">
"hmm!"
</Segment>
<Segment StartTime="1510.406" EndTime="1511.330" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1510.927" EndTime="1511.306" Participant="fe016">
"hmm" <Comment Description="shortened"/>
</Segment>
<Segment StartTime="1511.306" EndTime="1511.964" Participant="fe016">
"O_K", "uh" <Comment Description="shortened"/>
</Segment>
<Segment StartTime="1512.741" EndTime="1514.943" Participant="fe016">
Grunts, uh, <Emphasis> that </Emphasis> might be interesting.
</Segment>
<Segment StartTime="1513.356" EndTime="1514.642" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1514.945" EndTime="1515.487" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1515.256" EndTime="1515.774" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1515.487" EndTime="1516.845" Participant="fe008">
He's got lip - <Pause/> lipsmacks.
</Segment>
<Segment StartTime="1515.670" EndTime="1516.640" Participant="me001">
<VocalSound Description="clear throat"/>
</Segment>
<Segment StartTime="1515.760" EndTime="1516.643" Participant="fe016">
In the meetings.
</Segment>
<Segment StartTime="1516.845" EndTime="1517.959" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1516.856" EndTime="1517.573" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1516.898" EndTime="1517.356" Participant="me018">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1517.601" EndTime="1518.097" Participant="me018">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="1519.028" EndTime="1519.798" Participant="me013">
We should move on.
</Segment>
<Segment StartTime="1519.655" EndTime="1519.907" Participant="fe008">
Yeah.
</Segment>
<Segment StartTime="1519.912" EndTime="1520.371" Participant="me018">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="1520.549" EndTime="1521.278" Participant="me013">
Uh,
</Segment>
<Segment StartTime="1521.117" EndTime="1521.604" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1521.819" EndTime="1524.066" Participant="me013">
new version of, uh, presegmentation?
</Segment>
<Segment StartTime="1524.701" EndTime="1526.358" Participant="mn014">
Uh, oh yeah, um,
</Segment>
<Segment StartTime="1526.358" EndTime="1526.876" Participant="mn014">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="1526.876" EndTime="1535.543" Participant="mn014">
I worked a little bit on the - on the presegmentation to - to get another version which does channel-specific, uh, speech-nonspeech detection.
</Segment>
<Segment StartTime="1536.280" EndTime="1544.867" Participant="mn014">
And, what I did is I used some normalized features which, uh, look in- into the - which is normalized energy, uh,
</Segment>
<Segment StartTime="1544.867" EndTime="1548.328" Participant="mn014">
energy normalized by the mean over the channels and by the,
</Segment>
<Segment StartTime="1548.328" EndTime="1550.334" Participant="mn014">
minimum over the, <Uncertain> other </Uncertain>.
</Segment>
<Segment StartTime="1551.236" EndTime="1552.179" Participant="mn014">
within each channel.
</Segment>
<Segment StartTime="1552.891" EndTime="1554.533" Participant="mn014">
And to - to, mm,
</Segment>
<Segment StartTime="1555.041" EndTime="1561.761" Participant="mn014">
to, yeah, to normalize also loudness and - and modified loudness and things and that those special features <Uncertain> actually </Uncertain> are in my feature vector.
</Segment>
<Segment StartTime="1555.361" EndTime="1555.959" Participant="me011">
Oh.
</Segment>
<Segment StartTime="1562.500" EndTime="1563.695" Participant="mn014">
And, and,
</Segment>
<Segment StartTime="1563.695" EndTime="1571.587" Participant="mn014">
therefore to be able to, uh, somewhat distinguish between foreground and background speech in - in the different - in - each channel.
</Segment>
<Segment StartTime="1572.126" EndTime="1573.028" Participant="mn014">
And,
</Segment>
<Segment StartTime="1573.311" EndTime="1580.437" Participant="mn014">
eh, I tested it on - on three or four meetings and it seems to work, well yeah, fairly well, I - I would say.
</Segment>
<Segment StartTime="1580.828" EndTime="1582.348" Participant="mn014">
There are some problems with the lapel mike.
</Segment>
<Segment StartTime="1582.282" EndTime="1583.383" Participant="me011">
Of course. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1582.348" EndTime="1583.479" Participant="mn014">
<VocalSound Description="laugh"/> Yeah. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="1584.783" EndTime="1586.117" Participant="mn014">
Uh, yeah.
</Segment>
<Segment StartTime="1584.806" EndTime="1585.443" Participant="me011">
Wow that's great.
</Segment>
<Segment StartTime="1586.575" EndTime="1587.033" Participant="mn014">
And.
</Segment>
<Segment StartTime="1586.659" EndTime="1589.094" Participant="me011">
So I - I understand that's what you were saying about your problem with,
</Segment>
<Segment StartTime="1589.524" EndTime="1589.807" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1589.652" EndTime="1590.326" Participant="me011">
minimum.
</Segment>
<Segment StartTime="1589.807" EndTime="1590.090" Participant="mn014">
And.
</Segment>
<Segment StartTime="1590.310" EndTime="1593.799" Participant="mn014">
Yeah, and - and I had - I had, uh, specific problems with.
</Segment>
<Segment StartTime="1590.326" EndTime="1591.004" Participant="me011">
I get it.
</Segment>
<Segment StartTime="1592.040" EndTime="1594.806" Participant="me011">
So new- use ninetieth quartile, rather than,
</Segment>
<Segment StartTime="1594.294" EndTime="1594.631" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1594.954" EndTime="1595.331" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1594.986" EndTime="1595.542" Participant="me011">
minimum.
</Segment>
<Segment StartTime="1595.584" EndTime="1596.230" Participant="fe008">
Wow.
</Segment>
<Segment StartTime="1596.230" EndTime="1599.015" Participant="mn014">
Yeah - yeah, then - I - I did some - some - some things like that,
</Segment>
<Segment StartTime="1596.880" EndTime="1597.518" Participant="fe008">
Interesting.
</Segment>
<Segment StartTime="1599.284" EndTime="1601.385" Participant="mn014">
as there - there are some - some problems in,
</Segment>
<Segment StartTime="1601.668" EndTime="1602.287" Participant="mn014">
when,
</Segment>
<Segment StartTime="1602.839" EndTime="1606.813" Participant="mn014">
in the channel, there - <Uncertain> they </Uncertain> - <Uncertain> the </Uncertain> the speaker doesn't - doesn't talk much or doesn't talk at all.
</Segment>
<Segment StartTime="1607.285" EndTime="1608.087" Participant="mn014">
Then,
</Segment>
<Segment StartTime="1608.855" EndTime="1609.385" Participant="mn014">
the,
</Segment>
<Segment StartTime="1610.143" EndTime="1613.765" Participant="mn014">
yeah, there are - there are some problems with - with - with n- with normalization, and,
</Segment>
<Segment StartTime="1613.765" EndTime="1616.284" Participant="mn014">
then, uh, there the system doesn't work at all.
</Segment>
<Segment StartTime="1616.742" EndTime="1621.689" Participant="mn014">
So, I'm - I'm glad that there is the - the digit part, where everybody is forced to say something,
</Segment>
<Segment StartTime="1616.807" EndTime="1617.145" Participant="me013">
Right.
</Segment>
<Segment StartTime="1619.401" EndTime="1619.938" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1619.627" EndTime="1620.366" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1619.810" EndTime="1620.884" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1621.203" EndTime="1621.827" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1621.376" EndTime="1622.446" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1621.689" EndTime="1622.416" Participant="mn014">
so,
</Segment>
<Segment StartTime="1622.416" EndTime="1624.386" Participant="mn014">
that's - that's great for - for my purpose.
</Segment>
<Segment StartTime="1625.086" EndTime="1625.842" Participant="mn014">
And,
</Segment>
<Segment StartTime="1625.842" EndTime="1627.242" Participant="mn014">
the thing is I - I,
</Segment>
<Segment StartTime="1627.633" EndTime="1632.306" Participant="mn014">
then the evaluation of - of the system is a little bit hard, as I don't have any references.
</Segment>
<Segment StartTime="1631.993" EndTime="1633.833" Participant="me011">
Well we did the hand - the one by hand.
</Segment>
<Segment StartTime="1634.060" EndTime="1638.163" Participant="mn014">
Yeah, that's the one - one wh- where I do the training on so I can't do the evaluation on
</Segment>
<Segment StartTime="1637.160" EndTime="1638.759" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1637.960" EndTime="1640.470" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1638.029" EndTime="1638.704" Participant="me018">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1638.163" EndTime="1639.167" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1639.167" EndTime="1642.640" Participant="mn014">
So the thing is, can the transcribers perhaps do some,
</Segment>
<Segment StartTime="1640.470" EndTime="1641.028" Participant="me011">
Uh.
</Segment>
<Segment StartTime="1643.273" EndTime="1647.720" Participant="mn014">
some - some meetings in - in terms of speech-nonspeech in - in the specific channels?
</Segment>
<Segment StartTime="1647.610" EndTime="1649.683" Participant="me018">
Well won't you have that from their transcriptions?
</Segment>
<Segment StartTime="1647.953" EndTime="1648.670" Participant="fe008">
Well, I have -
</Segment>
<Segment StartTime="1649.487" EndTime="1652.488" Participant="fe008">
Well, O_K, so, now we need - so, um,
</Segment>
<Segment StartTime="1650.022" EndTime="1651.314" Participant="me011">
No, cuz we need is really tight.
</Segment>
<Segment StartTime="1651.044" EndTime="1651.361" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1652.488" EndTime="1656.851" Participant="fe008">
I think I might have done what you're requesting, though I did it in the service of a different thing.
</Segment>
<Segment StartTime="1655.401" EndTime="1656.182" Participant="mn014">
Oh, great.
</Segment>
<Segment StartTime="1656.851" EndTime="1661.952" Participant="fe008">
I have thirty minutes that I've more tightly transcribed with reference to individual channels.
</Segment>
<Segment StartTime="1660.829" EndTime="1661.315" Participant="mn014">
O_K.
</Segment>
<Segment StartTime="1661.991" EndTime="1662.841" Participant="mn014">
O_K, that's great.
</Segment>
<Segment StartTime="1662.841" EndTime="1664.171" Participant="mn014">
That's great for me. Yeah, so.
</Segment>
<Segment StartTime="1663.212" EndTime="1664.300" Participant="fe008">
And I could - And -
</Segment>
<Segment StartTime="1664.249" EndTime="1666.274" Participant="me011">
Hopefully that's not the same meeting that we did. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1664.329" EndTime="1664.825" Participant="fe008">
And -
</Segment>
<Segment StartTime="1666.120" EndTime="1667.515" Participant="fe008">
No, actually it's a different meeting.
</Segment>
<Segment StartTime="1667.108" EndTime="1667.469" Participant="me011">
Good.
</Segment>
<Segment StartTime="1667.515" EndTime="1670.245" Participant="fe008">
So, um, e- so the, you know, we have the,
</Segment>
<Segment StartTime="1667.971" EndTime="1668.281" Participant="mn014">
O_K.
</Segment>
<Segment StartTime="1670.245" EndTime="1681.343" Participant="fe008">
th- they transcribe as if it's one channel with these - with the slashes to separate the overlapping parts. And <Emphasis> then </Emphasis> we run it through - then it - then I'm gonna edit it and I'm gonna run it through channelize which takes it into Dave Gelbart's form- format. And then you have,
</Segment>
<Segment StartTime="1674.852" EndTime="1675.140" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1679.922" EndTime="1680.176" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1681.343" EndTime="1684.290" Participant="fe008">
all these things split across according to channel, and then that means that,
</Segment>
<Segment StartTime="1684.668" EndTime="1687.197" Participant="fe008">
if a person contributed more than once in a given,
</Segment>
<Segment StartTime="1687.197" EndTime="1689.209" Participant="fe008">
overlap during that time bend that -
</Segment>
<Segment StartTime="1689.522" EndTime="1692.690" Participant="fe008">
that two parts of the utterance end up together, it's the same channel,
</Segment>
<Segment StartTime="1692.690" EndTime="1696.197" Participant="fe008">
and then I took his tool, and last night for the first thirty minutes of
</Segment>
<Segment StartTime="1693.007" EndTime="1693.805" Participant="mn014">
O_K.
</Segment>
<Segment StartTime="1696.635" EndTime="1698.069" Participant="fe008">
one of these transcripts, I,
</Segment>
<Segment StartTime="1698.348" EndTime="1699.643" Participant="fe008">
tightened up the, um,
</Segment>
<Segment StartTime="1700.141" EndTime="1707.094" Participant="fe008">
boundaries on individual speakers' channels, cuz his - his interface <Emphasis> allows </Emphasis> me to have total flexibility in the time tags across the channels.
</Segment>
<Segment StartTime="1700.844" EndTime="1701.531" Participant="mn014">
O_K.
</Segment>
<Segment StartTime="1702.445" EndTime="1702.742" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1706.216" EndTime="1706.432" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1707.510" EndTime="1708.427" Participant="fe008">
And <Pause/> um,
</Segment>
<Segment StartTime="1709.047" EndTime="1709.705" Participant="me018">
So, current -
</Segment>
<Segment StartTime="1709.166" EndTime="1717.713" Participant="mn014">
so, yeah - yeah, that - that - that's great, but what <Emphasis> would </Emphasis> be nice to have some more meetings, not just one meeting to - to be sure that - that,
</Segment>
<Segment StartTime="1709.310" EndTime="1709.939" Participant="fe008">
so.
</Segment>
<Segment StartTime="1712.099" EndTime="1712.617" Participant="fe008">
Yes.
</Segment>
<Segment StartTime="1712.617" EndTime="1713.693" Participant="fe008">
Might not be what you need.
</Segment>
<Segment StartTime="1717.713" EndTime="1718.910" Participant="mn014">
there is a system,
</Segment>
<Segment StartTime="1718.089" EndTime="1722.380" Participant="me011">
Yeah, so if we could get a couple meetings done with that level of precision I think that would be a good idea.
</Segment>
<Segment StartTime="1720.163" EndTime="1720.809" Participant="mn014">
O_K.
</Segment>
<Segment StartTime="1722.380" EndTime="1722.847" Participant="fe008">
Oh, O_K.
</Segment>
<Segment StartTime="1722.587" EndTime="1722.829" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1723.465" EndTime="1731.294" Participant="fe008">
Uh, how - how m- much time - so the meetings vary in length, what are we talking about in terms of the number of minutes you'd like to have as your - as your training set?
</Segment>
<Segment StartTime="1732.640" EndTime="1735.382" Participant="mn014">
It seems to me <Uncertain> that </Uncertain> it <Emphasis> would </Emphasis> be good to have,
</Segment>
<Segment StartTime="1735.382" EndTime="1738.214" Participant="mn014">
a few minutes from - from different meetings, so.
</Segment>
<Segment StartTime="1738.982" EndTime="1741.040" Participant="mn014">
But I'm not sure about how much.
</Segment>
<Segment StartTime="1740.990" EndTime="1745.946" Participant="fe008">
O_K, now you're saying different meetings because of different <Emphasis> speakers </Emphasis> or because of different <Emphasis> audio </Emphasis> quality or <Emphasis> both </Emphasis> or - ? <Comment Description="leading question"/>
</Segment>
<Segment StartTime="1745.356" EndTime="1750.709" Participant="mn014">
Both - both. Different - different number of speakers, different speakers, different <Pause/> conditions.
</Segment>
<Segment StartTime="1745.946" EndTime="1746.464" Participant="fe008">
O_K.
</Segment>
<Segment StartTime="1749.690" EndTime="1753.865" Participant="me013">
Yeah, we don't have that much variety in meetings yet, uh, I mean we have this
</Segment>
<Segment StartTime="1752.574" EndTime="1752.992" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="1754.158" EndTime="1756.798" Participant="me013">
meeting and the feature meeting and we have a couple others that we have uh,
</Segment>
<Segment StartTime="1754.568" EndTime="1754.886" Participant="mn014">
Yeah, m-
</Segment>
<Segment StartTime="1756.798" EndTime="1757.976" Participant="me013">
couple examples of.
</Segment>
<Segment StartTime="1757.751" EndTime="1757.992" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1757.973" EndTime="1758.471" Participant="fe008">
Mm-hmm.
</Segment>
<Segment StartTime="1757.976" EndTime="1758.402" Participant="me013">
But -
</Segment>
<Segment StartTime="1758.402" EndTime="1759.400" Participant="me013">
but, uh,
</Segment>
<Segment StartTime="1759.651" EndTime="1763.269" Participant="me001">
Even probably with the gains <Pause/> differently will affect it, you mean -
</Segment>
<Segment StartTime="1763.723" EndTime="1767.220" Participant="mn014">
Uh, not really as - uh, because of the normalization, yeah. Yeah.
</Segment>
<Segment StartTime="1763.745" EndTime="1764.907" Participant="me013">
Poten- potentially.
</Segment>
<Segment StartTime="1764.803" EndTime="1766.842" Participant="me001">
Oh, cuz <Uncertain> you </Uncertain> use the normalization? O_K.
</Segment>
<Segment StartTime="1766.969" EndTime="1767.456" Participant="me013">
Oh, O_K.
</Segment>
<Segment StartTime="1767.371" EndTime="1769.959" Participant="fe016">
We can try running - we haven't done this yet because,
</Segment>
<Segment StartTime="1770.135" EndTime="1770.633" Participant="fe016">
um,
</Segment>
<Segment StartTime="1771.271" EndTime="1777.354" Participant="fe016">
uh, Andreas an- is - is gonna move over the S_R_I recognizer. i- basically I ran out of machines at S_R_I,
</Segment>
<Segment StartTime="1777.354" EndTime="1780.044" Participant="fe016">
cuz we're running the evals <Comment Description="meaning &quot;evaluations&quot;"/> and I just don't have machine time there.
</Segment>
<Segment StartTime="1777.596" EndTime="1777.967" Participant="mn014">
O_K.
</Segment>
<Segment StartTime="1780.440" EndTime="1782.009" Participant="fe016">
But, once that's moved over,
</Segment>
<Segment StartTime="1782.458" EndTime="1783.614" Participant="fe016">
uh, hopefully in a -
</Segment>
<Segment StartTime="1783.773" EndTime="1784.709" Participant="fe016">
a couple days,
</Segment>
<Segment StartTime="1785.068" EndTime="1786.616" Participant="fe016">
then, we can take,
</Segment>
<Segment StartTime="1786.856" EndTime="1787.413" Participant="fe016">
um,
</Segment>
<Segment StartTime="1788.091" EndTime="1790.380" Participant="fe016">
what Jane just told us about as,
</Segment>
<Segment StartTime="1788.800" EndTime="1790.135" Participant="me011">
Oh, shoot! <Comment Description="interjection"/>
</Segment>
<Segment StartTime="1790.559" EndTime="1792.054" Participant="fe016">
the presegmented,
</Segment>
<Segment StartTime="1792.503" EndTime="1792.970" Participant="fe016">
<VocalSound Description="breath"/> <NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="1792.970" EndTime="1795.113" Participant="fe016">
the - the segmentations that you did,
</Segment>
<Segment StartTime="1793.646" EndTime="1793.973" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1795.045" EndTime="1795.355" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1795.492" EndTime="1797.624" Participant="fe016">
at level eight or som- <Comment Description="while laughing"/> at some,
</Segment>
<Segment StartTime="1797.829" EndTime="1799.503" Participant="fe008">
The pre- presegment- <Pause/> yeah.
</Segment>
<Segment StartTime="1798.042" EndTime="1799.094" Participant="fe016">
threshold that Jane,
</Segment>
<Segment StartTime="1799.125" EndTime="1799.416" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1799.412" EndTime="1800.070" Participant="fe016">
tha- <Pause/> right,
</Segment>
<Segment StartTime="1800.707" EndTime="1801.876" Participant="fe016">
and try doing,
</Segment>
<Segment StartTime="1801.876" EndTime="1803.024" Participant="fe016">
forced alignment.
</Segment>
<Segment StartTime="1803.601" EndTime="1804.219" Participant="fe016">
um,
</Segment>
<Segment StartTime="1804.619" EndTime="1806.193" Participant="mn014">
With the recognizer? Yeah.
</Segment>
<Segment StartTime="1804.834" EndTime="1810.413" Participant="fe016">
on the word strings. And <Emphasis> if </Emphasis> it's good, then that will - that may give you a good boundary. Of course if it's good, we don't -
</Segment>
<Segment StartTime="1810.572" EndTime="1811.708" Participant="fe016">
then we're - we're fine, but,
</Segment>
<Segment StartTime="1810.677" EndTime="1811.365" Participant="mn014">
Yeah. M-
</Segment>
<Segment StartTime="1811.947" EndTime="1813.461" Participant="fe016">
I don't know yet whether these,
</Segment>
<Segment StartTime="1813.700" EndTime="1816.249" Participant="fe016">
segments that contain a lot of pauses around the words,
</Segment>
<Segment StartTime="1816.555" EndTime="1816.820" Participant="mn014">
I -
</Segment>
<Segment StartTime="1816.798" EndTime="1817.643" Participant="fe016">
will work or not.
</Segment>
<Segment StartTime="1817.653" EndTime="1828.233" Participant="mn014">
I would quite like to have some manually transcribed references for - for the system, as I'm not sure if - if it's really good to compare with - with some other automatic,
</Segment>
<Segment StartTime="1820.943" EndTime="1821.512" Participant="fe016">
Yeah.
</Segment>
<Segment StartTime="1828.290" EndTime="1828.693" Participant="fe016">
Right.
</Segment>
<Segment StartTime="1828.540" EndTime="1830.226" Participant="mn014">
found boundaries.
</Segment>
<Segment StartTime="1830.222" EndTime="1834.359" Participant="fe008">
Well, no, if we were to start with this and then tweak it h- manually, would that - that would be O_K?
</Segment>
<Segment StartTime="1831.912" EndTime="1832.310" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1833.162" EndTime="1833.601" Participant="fe016">
Right.
</Segment>
<Segment StartTime="1833.811" EndTime="1834.597" Participant="mn014">
Yeah <Pause/> sure.
</Segment>
<Segment StartTime="1834.204" EndTime="1837.471" Participant="fe016">
They might be O_K. It - you know it really depends on a lot of things, but,
</Segment>
<Segment StartTime="1834.757" EndTime="1835.056" Participant="fe008">
O_K.
</Segment>
<Segment StartTime="1836.067" EndTime="1836.364" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1837.471" EndTime="1839.244" Participant="fe016">
I would have maybe a transcriber,
</Segment>
<Segment StartTime="1839.563" EndTime="1844.542" Participant="fe016">
uh, look at the result of a forced alignment and then adjust those. That might save some time.
</Segment>
<Segment StartTime="1840.124" EndTime="1840.477" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1840.636" EndTime="1842.243" Participant="mn014">
To a- adjust them, or, yeah.
</Segment>
<Segment StartTime="1842.490" EndTime="1843.373" Participant="mn014">
Yeah, yeah.
</Segment>
<Segment StartTime="1844.863" EndTime="1847.837" Participant="fe016">
If they're horrible it won't help at all, but they might not be horrible.
</Segment>
<Segment StartTime="1845.142" EndTime="1845.800" Participant="mn014">
Yeah, great.
</Segment>
<Segment StartTime="1848.060" EndTime="1848.254" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1848.660" EndTime="1850.684" Participant="fe016">
So - but I'll let you know when we,
</Segment>
<Segment StartTime="1850.823" EndTime="1851.486" Participant="mn014">
O_K, great.
</Segment>
<Segment StartTime="1851.280" EndTime="1852.211" Participant="fe016">
uh, have that.
</Segment>
<Segment StartTime="1852.356" EndTime="1855.336" Participant="fe008">
How many minutes would you want from - I mean, we could <Pause/> easily,
</Segment>
<Segment StartTime="1855.336" EndTime="1861.160" Participant="fe008">
get a section, you know, like say a minute or so, from every meeting that we have so f- from the newer ones that we're working on,
</Segment>
<Segment StartTime="1861.698" EndTime="1862.780" Participant="fe008">
everyone that we have.
</Segment>
<Segment StartTime="1863.232" EndTime="1863.830" Participant="fe008">
And then,
</Segment>
<Segment StartTime="1866.179" EndTime="1867.255" Participant="fe008">
should provide this. <Comment Description="quietly"/>
</Segment>
<Segment StartTime="1866.893" EndTime="1872.088" Participant="mn014">
If it's not the <Emphasis> first </Emphasis> minute of - of the meeting, that - that's O_K with me, but, in - in the first minute, uh,
</Segment>
<Segment StartTime="1872.088" EndTime="1876.370" Participant="mn014">
Often there are some - some strange things going on which - which aren't really,
</Segment>
<Segment StartTime="1877.143" EndTime="1880.035" Participant="mn014">
well, for, which - which aren't re- re- really good.
</Segment>
<Segment StartTime="1880.406" EndTime="1880.724" Participant="mn014">
So.
</Segment>
<Segment StartTime="1881.333" EndTime="1883.193" Participant="mn014">
What - what I'd quite like, perhaps, is,
</Segment>
<Segment StartTime="1883.193" EndTime="1883.908" Participant="mn014">
to have,
</Segment>
<Segment StartTime="1884.509" EndTime="1887.833" Participant="mn014">
some five minutes of - of - of different <Pronounce Pronunciation="&quot;meterings&quot;"> meetings, </Pronounce> <Comment Description="rising"/> so.
</Segment>
<Segment StartTime="1888.689" EndTime="1891.700" Participant="fe008">
Somewhere <Emphasis> not </Emphasis> in the very beginning, five minutes, O_K.
</Segment>
<Segment StartTime="1891.445" EndTime="1891.719" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1892.357" EndTime="1894.966" Participant="fe008">
And, then I wanted to ask you just for my inter- information, then,
</Segment>
<Segment StartTime="1895.284" EndTime="1895.862" Participant="fe008">
would you,
</Segment>
<Segment StartTime="1896.181" EndTime="1896.978" Participant="fe008">
be trai-
</Segment>
<Segment StartTime="1897.396" EndTime="1899.861" Participant="fe008">
cuz I don't quite unders- so, would you be training then,
</Segment>
<Segment StartTime="1899.861" EndTime="1900.618" Participant="fe008">
um,
</Segment>
<Segment StartTime="1900.618" EndTime="1902.531" Participant="fe008">
the segmenter so that,
</Segment>
<Segment StartTime="1902.531" EndTime="1905.565" Participant="fe008">
it could, on the basis of that, segment the rest of the meeting?
</Segment>
<Segment StartTime="1905.565" EndTime="1909.016" Participant="fe008">
So, if I give you like <Pause/> five minutes is the idea that this would then be applied to,
</Segment>
<Segment StartTime="1909.723" EndTime="1910.472" Participant="fe008">
uh, to,
</Segment>
<Segment StartTime="1910.814" EndTime="1913.612" Participant="mn014">
I - I could do a - a retraining with that, yeah.
</Segment>
<Segment StartTime="1910.917" EndTime="1913.366" Participant="fe008">
providing tighter time <Pause/> bands?
</Segment>
<Segment StartTime="1913.612" EndTime="1914.636" Participant="mn014">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="1914.073" EndTime="1914.910" Participant="fe008">
Wow, interesting. <Comment Description="whispered"/>
</Segment>
<Segment StartTime="1914.636" EndTime="1917.959" Participant="mn014">
That's - but - but I hope that I - I don't <Emphasis> need </Emphasis> to do it.
</Segment>
<Segment StartTime="1915.298" EndTime="1915.702" Participant="fe008">
O_K.
</Segment>
<Segment StartTime="1918.444" EndTime="1921.431" Participant="mn014">
So, uh it c- can be do in an unsupervised way.
</Segment>
<Segment StartTime="1918.454" EndTime="1918.764" Participant="fe008">
Uh-huh.
</Segment>
<Segment StartTime="1922.903" EndTime="1924.026" Participant="fe008">
Excellent. <Emphasis> Excellent, </Emphasis> O_K.
</Segment>
<Segment StartTime="1922.968" EndTime="1923.312" Participant="mn014">
So.
</Segment>
<Segment StartTime="1924.407" EndTime="1925.299" Participant="mn014">
I'm - I'm not sure,
</Segment>
<Segment StartTime="1925.299" EndTime="1929.034" Participant="mn014">
but, for - for - for those three meetings whi- which I - which I did,
</Segment>
<Segment StartTime="1929.537" EndTime="1930.525" Participant="mn014">
it seems to be,
</Segment>
<Segment StartTime="1930.941" EndTime="1931.682" Participant="mn014">
quite well, but,
</Segment>
<Segment StartTime="1932.141" EndTime="1935.110" Participant="mn014">
there are some - some - as I said some problems with the lapel mike, but,
</Segment>
<Segment StartTime="1935.587" EndTime="1939.100" Participant="mn014">
perhaps we can do something with - with cross-correlations to,
</Segment>
<Segment StartTime="1939.263" EndTime="1940.690" Participant="mn014">
to get rid of the - of those.
</Segment>
<Segment StartTime="1941.352" EndTime="1941.815" Participant="mn014">
And.
</Segment>
<Segment StartTime="1943.109" EndTime="1950.243" Participant="mn014">
Yeah. That's - that's what I - that's my <Pause/> future work. Well - well what I want to do is to - to look into cross-correlations for - for removing those,
</Segment>
<Segment StartTime="1950.923" EndTime="1952.433" Participant="mn014">
false overlaps.
</Segment>
<Segment StartTime="1953.133" EndTime="1953.500" Participant="fe008">
Wonderful.
</Segment>
<Segment StartTime="1953.580" EndTime="1955.374" Participant="fe016">
Are the, um, wireless,
</Segment>
<Segment StartTime="1955.649" EndTime="1956.924" Participant="fe016">
different than the wired,
</Segment>
<Segment StartTime="1957.121" EndTime="1959.656" Participant="fe016">
mikes, at all? I mean, have you noticed any difference?
</Segment>
<Segment StartTime="1958.827" EndTime="1960.341" Participant="mn014">
I'm - I'm not sure, um,
</Segment>
<Segment StartTime="1960.620" EndTime="1970.119" Participant="mn014">
if - if there are any wired mikes in those meetings, or, uh, I have - have to loo- have a look at them but, I'm - I'm - I think there's no difference between,
</Segment>
<Segment StartTime="1970.050" EndTime="1972.244" Participant="fe016">
So it's just the lapel versus everything else?
</Segment>
<Segment StartTime="1971.115" EndTime="1971.613" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1972.012" EndTime="1972.291" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="1973.300" EndTime="1976.247" Participant="fe008">
O_K, so then, if that's five minutes per meeting we've got like twelve minutes,
</Segment>
<Segment StartTime="1976.601" EndTime="1977.460" Participant="fe008">
twelve meetings,
</Segment>
<Segment StartTime="1977.460" EndTime="1979.820" Participant="fe008">
roughly, that I'm - that I've been working with, then -
</Segment>
<Segment StartTime="1979.026" EndTime="1982.578" Participant="me013">
Of - of - of the meetings that you're working with, how many of them are different,
</Segment>
<Segment StartTime="1979.742" EndTime="1980.121" Participant="mn014">
No.
</Segment>
<Segment StartTime="1982.578" EndTime="1985.294" Participant="me013">
tha- are there any of them that are different than,
</Segment>
<Segment StartTime="1985.829" EndTime="1986.864" Participant="me013">
these two meetings?
</Segment>
<Segment StartTime="1986.983" EndTime="1989.887" Participant="fe008">
Well - oh wa- in terms of the speakers or the conditions or the?
</Segment>
<Segment StartTime="1989.311" EndTime="1991.154" Participant="me013">
Yeah, speakers. Sorry. So.
</Segment>
<Segment StartTime="1990.555" EndTime="1991.039" Participant="fe008">
Um,
</Segment>
<Segment StartTime="1990.630" EndTime="1991.840" Participant="mn014">
Yeah, that -
</Segment>
<Segment StartTime="1991.039" EndTime="1996.331" Participant="fe008">
we have different combinations of speakers. I mean, just from what I've seen, uh, there are some where,
</Segment>
<Segment StartTime="1996.728" EndTime="1997.199" Participant="fe008">
um,
</Segment>
<Segment StartTime="1997.199" EndTime="1998.892" Participant="fe008">
you're present or not present, and,
</Segment>
<Segment StartTime="1998.892" EndTime="2001.270" Participant="fe008">
then - then you have the difference between the networks group and this group
</Segment>
<Segment StartTime="2001.143" EndTime="2003.153" Participant="mn014">
Yeah, I know, some of the N_S_A meetings, yeah.
</Segment>
<Segment StartTime="2001.532" EndTime="2004.919" Participant="me013">
Yeah. So I didn't know in the group you had if you had - so you have the networks meeting?
</Segment>
<Segment StartTime="2003.153" EndTime="2003.613" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="2004.551" EndTime="2005.256" Participant="fe008">
Yep, we do.
</Segment>
<Segment StartTime="2004.588" EndTime="2004.927" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="2005.453" EndTime="2007.250" Participant="me013">
Do you have any of Jerry's meetings in your,
</Segment>
<Segment StartTime="2007.250" EndTime="2008.109" Participant="me013">
pack, er,
</Segment>
<Segment StartTime="2007.370" EndTime="2008.007" Participant="fe008">
Um, no.
</Segment>
<Segment StartTime="2008.109" EndTime="2008.559" Participant="me013">
No?
</Segment>
<Segment StartTime="2008.989" EndTime="2013.470" Participant="fe008">
We <Emphasis> could, </Emphasis> I mean you - you recorded one last week or so. I could get that new one in this week - I get that new one in.
</Segment>
<Segment StartTime="2011.255" EndTime="2011.594" Participant="me011">
Yep.
</Segment>
<Segment StartTime="2011.288" EndTime="2011.903" Participant="me018">
This week.
</Segment>
<Segment StartTime="2012.460" EndTime="2014.658" Participant="fe016">
We're gonna be recording them every <Pause/> Monday,
</Segment>
<Segment StartTime="2012.688" EndTime="2012.967" Participant="me011">
u-
</Segment>
<Segment StartTime="2013.865" EndTime="2014.317" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="2014.801" EndTime="2020.880" Participant="me013">
Cuz I think he really needs variety, and - and having as much variety for speaker certainly would be a big part of that I think.
</Segment>
<Segment StartTime="2015.044" EndTime="2015.582" Participant="fe016">
so -
</Segment>
<Segment StartTime="2016.640" EndTime="2016.953" Participant="fe008">
Great.
</Segment>
<Segment StartTime="2016.693" EndTime="2017.191" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="2020.778" EndTime="2021.097" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="2021.040" EndTime="2021.859" Participant="fe008">
O_K, so if I,
</Segment>
<Segment StartTime="2022.050" EndTime="2023.528" Participant="fe008">
O_K, included - include,
</Segment>
<Segment StartTime="2023.860" EndTime="2030.997" Participant="fe008">
O_K, then, uh, if I were to include all together samples from twelve meetings that would only take an hour and I could get the transcribers to do that right - I mean, what I mean is,
</Segment>
<Segment StartTime="2030.997" EndTime="2035.706" Participant="fe008">
that would be an hour <Emphasis> sampled, </Emphasis> and then they'd transcribe those - that hour, right? That's what I should do?
</Segment>
<Segment StartTime="2035.027" EndTime="2035.518" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="2036.867" EndTime="2037.271" Participant="me013">
And.
</Segment>
<Segment StartTime="2037.271" EndTime="2038.716" Participant="me013">
Right. Ye- But you're - y-
</Segment>
<Segment StartTime="2037.560" EndTime="2038.321" Participant="mn014">
That's - that's.
</Segment>
<Segment StartTime="2037.651" EndTime="2043.661" Participant="fe008">
I don't mean transcribe I mean - I mean adjust. So they get it into the multi-channel format and then adjust the timebands so it's precise.
</Segment>
<Segment StartTime="2042.540" EndTime="2045.630" Participant="me013">
So that should be faster than the ten times kind of thing, yeah.
</Segment>
<Segment StartTime="2044.494" EndTime="2046.408" Participant="fe008">
Absolutely. I did - I did, um,
</Segment>
<Segment StartTime="2046.964" EndTime="2047.622" Participant="fe008">
uh, so,
</Segment>
<Segment StartTime="2048.057" EndTime="2048.978" Participant="fe008">
last night I did,
</Segment>
<Segment StartTime="2049.622" EndTime="2049.966" Participant="fe008">
uh,
</Segment>
<Segment StartTime="2049.966" EndTime="2050.664" Participant="fe008">
Oh gosh, <Comment Description="whispered"/>
</Segment>
<Segment StartTime="2051.241" EndTime="2054.383" Participant="fe008">
well, last night, I did about half an hour in,
</Segment>
<Segment StartTime="2054.383" EndTime="2055.536" Participant="fe008">
three hours,
</Segment>
<Segment StartTime="2055.871" EndTime="2056.349" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="2056.315" EndTime="2056.926" Participant="fe008">
which is not,
</Segment>
<Segment StartTime="2057.574" EndTime="2058.777" Participant="fe008">
terrific, but, um,
</Segment>
<Segment StartTime="2058.990" EndTime="2060.194" Participant="me013">
Yeah. Well, that's probably.
</Segment>
<Segment StartTime="2059.111" EndTime="2060.821" Participant="fe008">
anyway, it's an hour and a half per -
</Segment>
<Segment StartTime="2062.078" EndTime="2062.456" Participant="mn014">
So.
</Segment>
<Segment StartTime="2062.187" EndTime="2062.429" Participant="fe008">
Well,
</Segment>
<Segment StartTime="2062.652" EndTime="2063.918" Participant="fe008">
I can't calculate on my,
</Segment>
<Segment StartTime="2064.130" EndTime="2069.650" Participant="mn014">
Do the transcribers actually start wi- with, uh, transcribing new meetings, or <Pause/> are they?
</Segment>
<Segment StartTime="2064.151" EndTime="2065.383" Participant="fe008">
<VocalSound Description="laugh"/> on my feet. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2069.381" EndTime="2074.523" Participant="fe008">
Well, um they're still working - they still have enough to finish that I haven't assigned a new meeting, but the next,
</Segment>
<Segment StartTime="2072.655" EndTime="2073.014" Participant="mn014">
O_K.
</Segment>
<Segment StartTime="2074.523" EndTime="2081.055" Participant="fe008">
m- m- I was about to need to assign a new meeting and I was going to take it from one of the new ones, and I could easily give them Jerry Feldman's meeting, no problem.
</Segment>
<Segment StartTime="2077.951" EndTime="2078.349" Participant="mn014">
O_K.
</Segment>
<Segment StartTime="2082.140" EndTime="2082.503" Participant="fe008">
And,
</Segment>
<Segment StartTime="2082.710" EndTime="2083.190" Participant="mn014">
O_K.
</Segment>
<Segment StartTime="2083.343" EndTime="2083.869" Participant="fe008">
then -
</Segment>
<Segment StartTime="2084.142" EndTime="2085.495" Participant="fe016">
So they're really running out of,
</Segment>
<Segment StartTime="2085.765" EndTime="2087.100" Participant="fe016">
data, prett- I mean that's good.
</Segment>
<Segment StartTime="2086.632" EndTime="2087.866" Participant="fe008">
Mm-hmm. Uh, that first set.
</Segment>
<Segment StartTime="2087.359" EndTime="2087.777" Participant="fe016">
Um,
</Segment>
<Segment StartTime="2087.777" EndTime="2088.295" Participant="fe016">
O_K.
</Segment>
<Segment StartTime="2088.523" EndTime="2091.833" Participant="me013">
They're running out of data unless we s- make the decision that we should go over and start,
</Segment>
<Segment StartTime="2089.172" EndTime="2089.926" Participant="fe016">
So -
</Segment>
<Segment StartTime="2092.543" EndTime="2093.931" Participant="me013">
uh, transcribing the other set.
</Segment>
<Segment StartTime="2094.376" EndTime="2096.025" Participant="me013">
There - the first - the first half.
</Segment>
<Segment StartTime="2100.379" EndTime="2103.316" Participant="fe008">
And so I was in the process of like editing them but this is <Emphasis> wonderful </Emphasis> news.
</Segment>
<Segment StartTime="2100.413" EndTime="2100.872" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="2102.944" EndTime="2103.462" Participant="mn014">
O_K.
</Segment>
<Segment StartTime="2103.609" EndTime="2103.890" Participant="me013">
Alright.
</Segment>
<Segment StartTime="2104.110" EndTime="2108.445" Participant="fe008">
We funded the experiment with, uh - also we were thinking maybe applying that that to getting the,
</Segment>
<Segment StartTime="2109.153" EndTime="2109.882" Participant="fe008">
Yeah, that'll be,
</Segment>
<Segment StartTime="2109.882" EndTime="2112.624" Participant="fe008">
very useful to getting the overlaps to be more precise all the way through.
</Segment>
<Segment StartTime="2113.445" EndTime="2116.402" Participant="me013">
So this, blends nicely into the update on transcripts.
</Segment>
<Segment StartTime="2116.300" EndTime="2116.872" Participant="fe008">
Yes, it does.
</Segment>
<Segment StartTime="2116.944" EndTime="2117.413" Participant="fe016">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="2116.988" EndTime="2117.306" Participant="me013">
<Pronounce Pronunciation="&quot;mm-kay&quot;"> O_K. </Pronounce>
</Segment>
<Segment StartTime="2117.277" EndTime="2117.904" Participant="fe008">
So, um, <Comment Description="mike noise"/>
</Segment>
<Segment StartTime="2118.461" EndTime="2118.967" Participant="fe008">
um,
</Segment>
<Segment StartTime="2119.183" EndTime="2120.378" Participant="fe016">
Yeah, please. Go ahead.
</Segment>
<Segment StartTime="2119.412" EndTime="2121.778" Participant="fe008">
Liz, and - and Don, and I met this morning,
</Segment>
<Segment StartTime="2122.021" EndTime="2123.357" Participant="fe008">
in the BARCO room,
</Segment>
<Segment StartTime="2123.522" EndTime="2124.625" Participant="fe008">
with the lecture hall, <Comment Description="very quiet"/>
</Segment>
<Segment StartTime="2123.582" EndTime="2125.457" Participant="fe016">
And this afternoon. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2124.901" EndTime="2125.917" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2125.241" EndTime="2127.467" Participant="fe008">
and this afternoon, it drifted into the afternoon, <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2125.457" EndTime="2126.334" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2126.832" EndTime="2127.270" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2127.467" EndTime="2127.973" Participant="fe008">
<VocalSound Description="inbreath"/>
</Segment>
<Segment StartTime="2127.973" EndTime="2129.683" Participant="fe008">
uh, concerning this issue of,
</Segment>
<Segment StartTime="2130.007" EndTime="2130.534" Participant="fe008">
um,
</Segment>
<Segment StartTime="2130.534" EndTime="2131.071" Participant="fe008">
the,
</Segment>
<Segment StartTime="2131.334" EndTime="2135.786" Participant="fe008">
well there's basically the issue of the interplay between the transcript format and the processing that,
</Segment>
<Segment StartTime="2136.139" EndTime="2137.149" Participant="fe008">
they need to do for,
</Segment>
<Segment StartTime="2137.361" EndTime="2138.586" Participant="fe008">
the S_R_I recognizer.
</Segment>
<Segment StartTime="2139.183" EndTime="2139.916" Participant="fe008">
And, um,
</Segment>
<Segment StartTime="2141.510" EndTime="2147.205" Participant="fe008">
well, so, I mentioned the process that I'm going through with the data, so, you know, I get the data back from the transcri- Well, s-
</Segment>
<Segment StartTime="2147.710" EndTime="2150.272" Participant="fe008">
uh, metaphorically, get the data back from the transcriber, and then I,
</Segment>
<Segment StartTime="2150.524" EndTime="2153.545" Participant="fe008">
check for simple things like spelling errors and things like that.
</Segment>
<Segment StartTime="2152.224" EndTime="2154.193" Participant="me018">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="2153.545" EndTime="2154.410" Participant="fe008">
And, um,
</Segment>
<Segment StartTime="2154.801" EndTime="2156.501" Participant="fe008">
I'm going to be doing a more thorough editing,
</Segment>
<Segment StartTime="2156.783" EndTime="2159.029" Participant="fe008">
with respect to consistency of the conventions.
</Segment>
<Segment StartTime="2159.505" EndTime="2161.326" Participant="fe008">
But they're - they're generally very good.
</Segment>
<Segment StartTime="2161.933" EndTime="2162.351" Participant="fe008">
And,
</Segment>
<Segment StartTime="2162.665" EndTime="2163.090" Participant="fe008">
then,
</Segment>
<Segment StartTime="2163.262" EndTime="2167.137" Participant="fe008">
I run it through, uh, the channelize program to get it into the multi-channel format,
</Segment>
<Segment StartTime="2168.154" EndTime="2168.691" Participant="fe008">
O_K.
</Segment>
<Segment StartTime="2169.039" EndTime="2169.911" Participant="fe008">
And <Pause/> the,
</Segment>
<Segment StartTime="2169.911" EndTime="2171.240" Participant="fe008">
what we discussed this morning,
</Segment>
<Segment StartTime="2171.432" EndTime="2173.138" Participant="fe008">
I would summarize as saying that,
</Segment>
<Segment StartTime="2173.575" EndTime="2173.981" Participant="fe008">
um,
</Segment>
<Segment StartTime="2174.416" EndTime="2176.177" Participant="fe008">
these units that result,
</Segment>
<Segment StartTime="2176.177" EndTime="2178.767" Participant="fe008">
in a - a particular channel and a particular timeband,
</Segment>
<Segment StartTime="2178.939" EndTime="2179.991" Participant="fe008">
at - at that level,
</Segment>
<Segment StartTime="2180.446" EndTime="2180.801" Participant="fe008">
um,
</Segment>
<Segment StartTime="2181.124" EndTime="2182.298" Participant="fe008">
vary in length.
</Segment>
<Segment StartTime="2182.621" EndTime="2183.096" Participant="fe008">
And,
</Segment>
<Segment StartTime="2183.096" EndTime="2183.551" Participant="fe008">
um,
</Segment>
<Segment StartTime="2183.551" EndTime="2183.865" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2183.865" EndTime="2187.172" Participant="fe008">
their recognizer would prefer that the units not be overly long.
</Segment>
<Segment StartTime="2187.375" EndTime="2189.014" Participant="fe008">
But it's really an empirical question,
</Segment>
<Segment StartTime="2189.156" EndTime="2191.676" Participant="fe008">
whether the units we get at this point through,
</Segment>
<Segment StartTime="2191.676" EndTime="2194.831" Participant="fe008">
just that process I described might be sufficient for them. So,
</Segment>
<Segment StartTime="2195.034" EndTime="2199.264" Participant="fe008">
as a first pass through, a first chance without having to do a lot of hand-editing,
</Segment>
<Segment StartTime="2199.618" EndTime="2200.457" Participant="fe008">
what we're gonna do,
</Segment>
<Segment StartTime="2200.643" EndTime="2202.379" Participant="fe008">
is, I'll run it through channelize,
</Segment>
<Segment StartTime="2202.379" EndTime="2205.458" Participant="fe008">
give them those data after I've done the editing process and be sure it's clean.
</Segment>
<Segment StartTime="2205.458" EndTime="2205.954" Participant="fe008">
And I can do that,
</Segment>
<Segment StartTime="2206.177" EndTime="2206.819" Participant="fe008">
pretty quickly,
</Segment>
<Segment StartTime="2207.250" EndTime="2207.740" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="2207.294" EndTime="2207.876" Participant="fe008">
with just,
</Segment>
<Segment StartTime="2207.876" EndTime="2208.950" Participant="fe008">
that minimal editing,
</Segment>
<Segment StartTime="2209.270" EndTime="2210.549" Participant="fe008">
without having to hand-break things.
</Segment>
<Segment StartTime="2210.766" EndTime="2212.565" Participant="fe008">
And then we'll see if the units that we're getting,
</Segment>
<Segment StartTime="2213.150" EndTime="2213.480" Participant="fe008">
uh,
</Segment>
<Segment StartTime="2213.872" EndTime="2215.059" Participant="fe008">
with the - at that level,
</Segment>
<Segment StartTime="2215.059" EndTime="2215.956" Participant="fe008">
are sufficient.
</Segment>
<Segment StartTime="2215.956" EndTime="2218.309" Participant="fe008">
And maybe they don't <Emphasis> need </Emphasis> to be further broken down.
</Segment>
<Segment StartTime="2218.623" EndTime="2222.648" Participant="fe008">
And if they do need to be further broken down then maybe it just be piece-wise, maybe it won't be the whole thing.
</Segment>
<Segment StartTime="2223.236" EndTime="2225.507" Participant="fe008">
So, that's - that's what we were discussing,
</Segment>
<Segment StartTime="2226.508" EndTime="2227.219" Participant="fe008">
this morning as far as I -
</Segment>
<Segment StartTime="2226.540" EndTime="2227.053" Participant="fe016">
Right.
</Segment>
<Segment StartTime="2227.363" EndTime="2230.284" Participant="fe008">
Among - also we discussed some adaptational things, so it's like,
</Segment>
<Segment StartTime="2228.520" EndTime="2229.384" Participant="fe016">
Then lots of -
</Segment>
<Segment StartTime="2229.663" EndTime="2230.141" Participant="fe016">
Right.
</Segment>
<Segment StartTime="2230.284" EndTime="2230.625" Participant="fe008">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="2230.625" EndTime="2231.058" Participant="fe008">
uh -
</Segment>
<Segment StartTime="2231.480" EndTime="2233.513" Participant="fe008">
You know I hadn't, uh, incorporated,
</Segment>
<Segment StartTime="2233.513" EndTime="2236.599" Participant="fe008">
a convention explicitly to handle acronyms, for example,
</Segment>
<Segment StartTime="2236.599" EndTime="2237.631" Participant="fe008">
but if someone says,
</Segment>
<Segment StartTime="2237.951" EndTime="2238.714" Participant="fe008">
P_Z_M
</Segment>
<Segment StartTime="2238.974" EndTime="2242.380" Participant="fe008">
it would be nice to have that be directly interpretable from,
</Segment>
<Segment StartTime="2242.380" EndTime="2245.920" Participant="fe008">
the transcript what they said, or Pi- uh <Pronounce Pronunciation="&quot;tickle&quot;"> Tcl </Pronounce> - T_C_L I mean. It's like
</Segment>
<Segment StartTime="2242.440" EndTime="2242.943" Participant="me013">
Mm-hmm.
</Segment>
<Segment StartTime="2245.920" EndTime="2247.034" Participant="fe008">
y- it's - and so,
</Segment>
<Segment StartTime="2247.457" EndTime="2247.910" Participant="fe008">
um,
</Segment>
<Segment StartTime="2248.540" EndTime="2250.717" Participant="fe008">
I've - I've incorporated also convention,
</Segment>
<Segment StartTime="2250.975" EndTime="2254.391" Participant="fe008">
with that but that's easy to handle at the post editing phase, and I'll mention it to,
</Segment>
<Segment StartTime="2254.391" EndTime="2256.517" Participant="fe008">
transcribers for the next phase but that's O_K. And then,
</Segment>
<Segment StartTime="2256.981" EndTime="2261.473" Participant="fe008">
a similar conv- uh, convention for <Emphasis> numbers. </Emphasis> So if they say one-eighty-three versus one eight three.
</Segment>
<Segment StartTime="2262.223" EndTime="2262.563" Participant="fe008">
Um,
</Segment>
<Segment StartTime="2262.738" EndTime="2264.079" Participant="fe008">
and also I'll be, um,
</Segment>
<Segment StartTime="2264.740" EndTime="2267.525" Participant="fe008">
encoding, as I do my post-editing, the,
</Segment>
<Segment StartTime="2267.866" EndTime="2270.673" Participant="fe008">
things that are in curly brackets, which are clarificational material.
</Segment>
<Segment StartTime="2270.911" EndTime="2273.687" Participant="fe008">
And eh to incorporate, uh, keyword,
</Segment>
<Segment StartTime="2273.687" EndTime="2274.754" Participant="fe008">
at the beginning. So,
</Segment>
<Segment StartTime="2274.754" EndTime="2277.561" Participant="fe008">
it's gonna be either a gloss or it's gonna be a vocal sound like a,
</Segment>
<Segment StartTime="2277.788" EndTime="2278.829" Participant="fe008">
laugh or a cough,
</Segment>
<Segment StartTime="2279.211" EndTime="2279.975" Participant="fe008">
or, so forth.
</Segment>
<Segment StartTime="2280.243" EndTime="2283.906" Participant="fe008">
Or a non-vocal sound like a doors- door-slam, and that can be easily done with a,
</Segment>
<Segment StartTime="2284.732" EndTime="2287.116" Participant="fe008">
you know, just a - one little additional thing in the,
</Segment>
<Segment StartTime="2288.075" EndTime="2289.035" Participant="fe008">
in the general format.
</Segment>
<Segment StartTime="2289.682" EndTime="2291.722" Participant="fe016">
Yeah we j- we just needed a way to,
</Segment>
<Segment StartTime="2292.120" EndTime="2295.067" Participant="fe016">
strip, you know, all the comments, all the things th- the - that
</Segment>
<Segment StartTime="2295.067" EndTime="2297.757" Participant="fe016">
linguist wants but the recognizer can't do anything with.
</Segment>
<Segment StartTime="2298.514" EndTime="2299.072" Participant="fe016">
Um,
</Segment>
<Segment StartTime="2299.072" EndTime="2302.638" Participant="fe016">
but to keep things that we mapped to like reject models, or, you know,
</Segment>
<Segment StartTime="2302.638" EndTime="2304.215" Participant="fe016">
uh, mouth noise, or,
</Segment>
<Segment StartTime="2304.631" EndTime="2305.420" Participant="fe016">
cough.
</Segment>
<Segment StartTime="2305.938" EndTime="2310.003" Participant="fe016">
And then there's this interesting issue Jane brought up which I hadn't thought about before but I was,
</Segment>
<Segment StartTime="2310.003" EndTime="2313.230" Participant="fe016">
realizing as I went through the transcripts, that there are some noises like,
</Segment>
<Segment StartTime="2314.207" EndTime="2318.483" Participant="fe016">
um, well the - good example was an inbreath, where a transcriber working from,
</Segment>
<Segment StartTime="2318.822" EndTime="2319.635" Participant="fe016">
the mixed,
</Segment>
<Segment StartTime="2319.735" EndTime="2320.516" Participant="fe016">
signal,
</Segment>
<Segment StartTime="2320.611" EndTime="2325.353" Participant="fe016">
doesn't know whose breath it is, and they've been assigning it to someone that may or may not be correct.
</Segment>
<Segment StartTime="2322.125" EndTime="2322.663" Participant="me011">
Right.
</Segment>
<Segment StartTime="2325.791" EndTime="2329.894" Participant="fe016">
And what we do is, if it's a breath sound, you know, a sound from the speaker, we map it,
</Segment>
<Segment StartTime="2330.412" EndTime="2330.982" Participant="fe016">
to,
</Segment>
<Segment StartTime="2331.338" EndTime="2334.463" Participant="fe016">
a noise model, like a mouth-noise model in the recognizer, and,
</Segment>
<Segment StartTime="2334.463" EndTime="2337.469" Participant="fe016">
yeah, it probably doesn't hurt that much once in a while to have these, but,
</Segment>
<Segment StartTime="2337.971" EndTime="2339.665" Participant="fe016">
if they're in the wrong channel, that's,
</Segment>
<Segment StartTime="2339.665" EndTime="2340.679" Participant="fe016">
not a good idea.
</Segment>
<Segment StartTime="2340.998" EndTime="2342.125" Participant="fe016">
And then there's also,
</Segment>
<Segment StartTime="2342.532" EndTime="2345.350" Participant="fe016">
things like door-slams that's really in no one's channel, they're like -
</Segment>
<Segment StartTime="2345.529" EndTime="2346.506" Participant="fe016">
it's in the room. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2346.567" EndTime="2347.104" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="2346.581" EndTime="2347.049" Participant="me011">
Right.
</Segment>
<Segment StartTime="2347.143" EndTime="2350.132" Participant="fe016">
And <Pause/> uh, Jane had this nice, uh,
</Segment>
<Segment StartTime="2350.351" EndTime="2351.427" Participant="fe016">
idea of having,
</Segment>
<Segment StartTime="2351.427" EndTime="2352.304" Participant="fe016">
like an extra,
</Segment>
<Segment StartTime="2352.872" EndTime="2353.849" Participant="me011">
An extra channel.
</Segment>
<Segment StartTime="2353.021" EndTime="2354.664" Participant="fe016">
uh couple tiers, yeah.
</Segment>
<Segment StartTime="2354.090" EndTime="2356.458" Participant="fe008">
Yeah. I've been - I've been adding that to the ones I've been editing.
</Segment>
<Segment StartTime="2355.282" EndTime="2356.617" Participant="fe016">
And we were thinking,
</Segment>
<Segment StartTime="2356.617" EndTime="2365.821" Participant="fe016">
that is useful also when there's uncertainties. So if they hear a breath and they don't know who breath it is it's better to put it in that channel than to put it in the speaker's channel because maybe it was someone else's breath, or -
</Segment>
<Segment StartTime="2360.733" EndTime="2361.019" Participant="me013">
<NonVocalSound Description="door slams"/>
</Segment>
<Segment StartTime="2361.019" EndTime="2362.411" Participant="me013">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="2366.592" EndTime="2370.094" Participant="fe016">
Uh, so I think that's a good - you can always clean that up, post-processing.
</Segment>
<Segment StartTime="2370.514" EndTime="2370.741" Participant="fe008">
Yeah.
</Segment>
<Segment StartTime="2370.532" EndTime="2372.724" Participant="fe016">
So a lot of little details, but I think we're,
</Segment>
<Segment StartTime="2370.566" EndTime="2373.501" Participant="me013">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="2373.500" EndTime="2375.215" Participant="fe016">
coming to some kinda closure,
</Segment>
<Segment StartTime="2376.308" EndTime="2378.121" Participant="fe016">
on that. So the idea is then,
</Segment>
<Segment StartTime="2378.420" EndTime="2379.656" Participant="fe016">
uh, Don can take,
</Segment>
<Segment StartTime="2379.875" EndTime="2382.126" Participant="fe016">
uh, Jane's post-processed channelized version,
</Segment>
<Segment StartTime="2382.525" EndTime="2383.141" Participant="fe016">
and,
</Segment>
<Segment StartTime="2383.141" EndTime="2385.492" Participant="fe016">
with some scripts, you know, convert that to -
</Segment>
<Segment StartTime="2385.731" EndTime="2387.601" Participant="fe016">
to a reference for the recognizer and we can,
</Segment>
<Segment StartTime="2386.949" EndTime="2387.348" Participant="me018">
<VocalSound Description="swallow"/>
</Segment>
<Segment StartTime="2388.201" EndTime="2388.960" Participant="fe016">
can run these.
</Segment>
<Segment StartTime="2388.960" EndTime="2390.250" Participant="fe016">
So <Pause/> when that's,
</Segment>
<Segment StartTime="2390.250" EndTime="2394.474" Participant="fe016">
ready - you know, as soon as that's ready, and as soon as the recognizer is here we can get,
</Segment>
<Segment StartTime="2395.131" EndTime="2398.623" Participant="fe016">
twelve hours of force-aligned and recognized data.
</Segment>
<Segment StartTime="2399.400" EndTime="2400.321" Participant="fe016">
And, you know,
</Segment>
<Segment StartTime="2400.541" EndTime="2401.019" Participant="fe016">
start,
</Segment>
<Segment StartTime="2401.357" EndTime="2402.314" Participant="fe016">
working on it, so we're,
</Segment>
<Segment StartTime="2401.427" EndTime="2401.768" Participant="fe008">
And -
</Segment>
<Segment StartTime="2402.533" EndTime="2405.223" Participant="fe016">
I dunno a coup- a week or two away I would say from,
</Segment>
<Segment StartTime="2406.099" EndTime="2409.624" Participant="fe016">
uh, if - if that process is automatic once we get your post-process,
</Segment>
<Segment StartTime="2410.002" EndTime="2410.809" Participant="fe016">
transcript.
</Segment>
<Segment StartTime="2411.930" EndTime="2412.249" Participant="fe008">
Mm-hmm.
</Segment>
<Segment StartTime="2412.249" EndTime="2417.813" Participant="fe008">
And that doesn't - the amount of editing that it would require is not very much either. I'm just <Emphasis> hoping </Emphasis> that the units that are provided in that way,
</Segment>
<Segment StartTime="2417.813" EndTime="2418.050" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2418.050" EndTime="2420.292" Participant="fe008">
will be sufficient cuz I would save a lot of, uh,
</Segment>
<Segment StartTime="2420.847" EndTime="2421.174" Participant="fe008">
time,
</Segment>
<Segment StartTime="2421.937" EndTime="2422.957" Participant="fe008">
dividing things.
</Segment>
<Segment StartTime="2423.508" EndTime="2425.171" Participant="fe016">
Yeah, some of them are quite long.
</Segment>
<Segment StartTime="2425.760" EndTime="2428.109" Participant="fe016">
Just from - I dunno how long were - you did one?
</Segment>
<Segment StartTime="2427.368" EndTime="2427.588" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2428.130" EndTime="2431.194" Participant="me001">
I saw a couple, <VocalSound Description="clear throat"/> around twenty seconds,
</Segment>
<Segment StartTime="2431.194" EndTime="2434.103" Participant="me001">
and that was just without looking too hard for it, so,
</Segment>
<Segment StartTime="2433.760" EndTime="2434.206" Participant="fe016">
Right.
</Segment>
<Segment StartTime="2434.200" EndTime="2435.002" Participant="fe008">
Well n-
</Segment>
<Segment StartTime="2434.601" EndTime="2436.593" Participant="me001">
I would imagine that there might be some that are longer.
</Segment>
<Segment StartTime="2436.604" EndTime="2440.300" Participant="fe008">
One question, e- w- would that be a single speaker or is that multiple speakers overlapping?
</Segment>
<Segment StartTime="2438.904" EndTime="2439.504" Participant="me001">
No.
</Segment>
<Segment StartTime="2440.022" EndTime="2443.310" Participant="me001">
No, but if we're gonna segment it, like if there's one speaker in there,
</Segment>
<Segment StartTime="2443.530" EndTime="2445.104" Participant="me001">
that says "O_K" or something,
</Segment>
<Segment StartTime="2445.502" EndTime="2448.708" Participant="me001">
right in the middle, it's gonna have a lot of dead time <Emphasis> around </Emphasis> it, so it's not -
</Segment>
<Segment StartTime="2446.924" EndTime="2448.308" Participant="fe016">
Right. It's not the -
</Segment>
<Segment StartTime="2448.704" EndTime="2450.617" Participant="me018">
<VocalSound Description="yawn"/>
</Segment>
<Segment StartTime="2448.832" EndTime="2452.562" Participant="fe016">
it's not the fact that we can't process a twenty second segment, it's the fact that,
</Segment>
<Segment StartTime="2452.860" EndTime="2456.467" Participant="fe016">
there's twenty seconds in which to place one word in the wrong place- <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2456.336" EndTime="2457.093" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2456.837" EndTime="2457.522" Participant="me001">
Yeah. <Comment Description="whispered"/>
</Segment>
<Segment StartTime="2456.965" EndTime="2457.662" Participant="fe016">
You know, if -
</Segment>
<Segment StartTime="2457.144" EndTime="2457.540" Participant="fe008">
Yeah.
</Segment>
<Segment StartTime="2457.662" EndTime="2460.471" Participant="fe016">
if someone has a very short utterance there, and that's where,
</Segment>
<Segment StartTime="2460.879" EndTime="2461.337" Participant="fe016">
we,
</Segment>
<Segment StartTime="2461.556" EndTime="2463.788" Participant="fe016">
might wanna have this individual, you know,
</Segment>
<Segment StartTime="2463.413" EndTime="2463.891" Participant="mn014">
Yep.
</Segment>
<Segment StartTime="2463.788" EndTime="2465.826" Participant="fe016">
ha- have your pre- pre-process input.
</Segment>
<Segment StartTime="2465.106" EndTime="2465.485" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="2465.485" EndTime="2465.855" Participant="mn014">
Sure.
</Segment>
<Segment StartTime="2465.855" EndTime="2471.721" Participant="mn014">
I - I - I thought that perhaps the transcribers could start then from the - those mult- multi-channel,
</Segment>
<Segment StartTime="2466.059" EndTime="2467.471" Participant="fe016">
And I just don't know, I have to run it.
</Segment>
<Segment StartTime="2466.120" EndTime="2466.989" Participant="fe008">
That's very important.
</Segment>
<Segment StartTime="2472.151" EndTime="2475.456" Participant="mn014">
uh, speech-nonspeech detections, if they would like to.
</Segment>
<Segment StartTime="2474.427" EndTime="2474.944" Participant="fe016">
Right.
</Segment>
<Segment StartTime="2474.910" EndTime="2477.276" Participant="fe008">
In - in doing the hand-marking? Yeah that's what I was thinking, too.
</Segment>
<Segment StartTime="2475.845" EndTime="2476.203" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="2476.808" EndTime="2477.206" Participant="fe016">
Right.
</Segment>
<Segment StartTime="2477.405" EndTime="2480.434" Participant="fe016">
So that's probably what will happen, but we'll try it this way and see.
</Segment>
<Segment StartTime="2477.585" EndTime="2477.853" Participant="fe008">
Yeah.
</Segment>
<Segment StartTime="2479.840" EndTime="2480.258" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="2480.992" EndTime="2483.074" Participant="fe016">
I mean it's probably good enough for force-alignment.
</Segment>
<Segment StartTime="2483.572" EndTime="2484.863" Participant="fe016">
If it's not then we're really -
</Segment>
<Segment StartTime="2484.863" EndTime="2485.939" Participant="fe016">
then we def- definitely
</Segment>
<Segment StartTime="2485.279" EndTime="2485.678" Participant="mn014">
Yeah.
</Segment>
<Segment StartTime="2486.198" EndTime="2491.438" Participant="fe016">
uh, but for free recognition I'm - it'll probably not be good enough. We'll probably get lots of errors because of the cross-talk, and,
</Segment>
<Segment StartTime="2491.752" EndTime="2492.151" Participant="mn014">
Yep.
</Segment>
<Segment StartTime="2491.996" EndTime="2492.950" Participant="fe016">
noises and things.
</Segment>
<Segment StartTime="2496.396" EndTime="2498.644" Participant="me013">
Good <Pause/> s- I think that's
</Segment>
<Segment StartTime="2499.241" EndTime="2501.436" Participant="me013">
probably our agenda, or starting up there.
</Segment>
<Segment StartTime="2500.131" EndTime="2501.588" Participant="fe008">
Oh I wanted to ask one thing,
</Segment>
<Segment StartTime="2501.436" EndTime="2501.948" Participant="me013">
Yeah?
</Segment>
<Segment StartTime="2501.875" EndTime="2504.104" Participant="fe008">
the microphones - the new microphones, when do we get, uh?
</Segment>
<Segment StartTime="2504.176" EndTime="2505.926" Participant="me011">
Uh, they said it would take about a week.
</Segment>
<Segment StartTime="2505.948" EndTime="2506.602" Participant="fe008">
Oh, exciting.
</Segment>
<Segment StartTime="2506.830" EndTime="2507.028" Participant="fe008">
K_.
</Segment>
<Segment StartTime="2506.935" EndTime="2507.410" Participant="me013">
K_.
</Segment>
<Segment StartTime="2506.975" EndTime="2507.810" Participant="me018">
You ordered <Pronounce Pronunciation="&quot;em&quot;"> them </Pronounce> already?
</Segment>
<Segment StartTime="2507.452" EndTime="2507.850" Participant="me011">
Mm-hmm.
</Segment>
<Segment StartTime="2508.208" EndTime="2508.706" Participant="me018">
Great.
</Segment>
<Segment StartTime="2508.955" EndTime="2511.253" Participant="fe016">
So what happens to our old microphones?
</Segment>
<Segment StartTime="2512.260" EndTime="2514.279" Participant="me013">
They go where old microphones go.
</Segment>
<Segment StartTime="2512.452" EndTime="2513.050" Participant="me011">
Um -
</Segment>
<Segment StartTime="2512.719" EndTime="2514.846" Participant="fe016">
Do we give them to someone, <Pause/> or - ? <Comment Description="leading question"/>
</Segment>
<Segment StartTime="2514.279" EndTime="2515.914" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2515.151" EndTime="2516.083" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2515.301" EndTime="2516.393" Participant="me018">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2515.572" EndTime="2517.433" Participant="me011">
Well the only thing we're gonna have extra,
</Segment>
<Segment StartTime="2518.332" EndTime="2520.284" Participant="fe016">
We don't have more <Emphasis> receivers, </Emphasis> we just have -
</Segment>
<Segment StartTime="2518.429" EndTime="2519.433" Participant="me011">
for now,
</Segment>
<Segment StartTime="2520.131" EndTime="2523.132" Participant="me011">
Right, we don- so the only thing we'll have extra now is just the lapel.
</Segment>
<Segment StartTime="2520.782" EndTime="2521.294" Participant="fe016">
Right.
</Segment>
<Segment StartTime="2523.411" EndTime="2524.347" Participant="me011">
Not - not the,
</Segment>
<Segment StartTime="2524.347" EndTime="2525.164" Participant="me011">
<Emphasis> bodypack, </Emphasis>
</Segment>
<Segment StartTime="2525.383" EndTime="2526.499" Participant="me011">
just the lapel.
</Segment>
<Segment StartTime="2525.828" EndTime="2527.090" Participant="fe016">
Just the lapel itself.
</Segment>
<Segment StartTime="2526.904" EndTime="2529.470" Participant="me011">
Um, and then one of the - one of those.
</Segment>
<Segment StartTime="2529.713" EndTime="2533.518" Participant="me011">
Since, what I decided to do, on Morgan's suggestion, was just get two,
</Segment>
<Segment StartTime="2533.518" EndTime="2534.754" Participant="me011">
new microphones,
</Segment>
<Segment StartTime="2535.153" EndTime="2536.293" Participant="me011">
um, and try them out.
</Segment>
<Segment StartTime="2536.390" EndTime="2537.160" Participant="fe016">
Mm-hmm.
</Segment>
<Segment StartTime="2536.672" EndTime="2538.324" Participant="me011">
And then, if we <Emphasis> like </Emphasis> them we'll get more.
</Segment>
<Segment StartTime="2538.699" EndTime="2539.197" Participant="fe016">
O_K.
</Segment>
<Segment StartTime="2538.997" EndTime="2539.402" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="2539.370" EndTime="2541.869" Participant="me011">
Since they're - they're like two hundred bucks a piece, we won't, uh,
</Segment>
<Segment StartTime="2542.965" EndTime="2543.981" Participant="me011">
at least try <Pronounce Pronunciation="&quot;em&quot;"> them </Pronounce> out.
</Segment>
<Segment StartTime="2543.642" EndTime="2545.894" Participant="me018">
So it's a replacement for this headset mike?
</Segment>
<Segment StartTime="2544.937" EndTime="2545.355" Participant="me011">
Yep.
</Segment>
<Segment StartTime="2545.913" EndTime="2546.212" Participant="me011">
Yep.
</Segment>
<Segment StartTime="2546.008" EndTime="2546.535" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="2547.268" EndTime="2549.022" Participant="me011">
And they're gonna do the wiring for us.
</Segment>
<Segment StartTime="2549.132" EndTime="2551.065" Participant="me018">
What's the, um, style of the headset?
</Segment>
<Segment StartTime="2551.034" EndTime="2552.110" Participant="me011">
It's, um,
</Segment>
<Segment StartTime="2552.430" EndTime="2556.555" Participant="me011">
it's by Crown, and it's one of these sort of mount around the ear thingies,
</Segment>
<Segment StartTime="2556.933" EndTime="2557.650" Participant="me011">
and,
</Segment>
<Segment StartTime="2557.650" EndTime="2562.472" Participant="me011">
uh, when I s- when I <Emphasis> mentioned </Emphasis> that we thought it was uncomfortable he said it was a common problem with the Sony.
</Segment>
<Segment StartTime="2562.751" EndTime="2565.205" Participant="me011">
And this is how apparently a lot of people are getting around it.
</Segment>
<Segment StartTime="2565.223" EndTime="2565.809" Participant="me018">
Hmm.
</Segment>
<Segment StartTime="2565.564" EndTime="2568.661" Participant="me011">
And I checked on the web, and <Emphasis> every </Emphasis> site I went to,
</Segment>
<Segment StartTime="2568.880" EndTime="2570.633" Participant="me011">
raved about this particular mike.
</Segment>
<Segment StartTime="2571.111" EndTime="2574.493" Participant="me011">
It's apparently comfortable and stays on the head well, so we'll see if it's any good.
</Segment>
<Segment StartTime="2574.870" EndTime="2575.649" Participant="me011">
But, uh,
</Segment>
<Segment StartTime="2576.383" EndTime="2577.498" Participant="me011">
I think it's promising.
</Segment>
<Segment StartTime="2578.180" EndTime="2580.612" Participant="fe008">
You said it was used by aerobics instructors?
</Segment>
<Segment StartTime="2579.895" EndTime="2580.253" Participant="me011">
Yep.
</Segment>
<Segment StartTime="2580.552" EndTime="2581.654" Participant="me011">
Yep, so it was -
</Segment>
<Segment StartTime="2580.760" EndTime="2581.183" Participant="me001">
Hmm.
</Segment>
<Segment StartTime="2580.899" EndTime="2581.672" Participant="fe008">
That says a lot.
</Segment>
<Segment StartTime="2581.672" EndTime="2583.316" Participant="fe008">
<VocalSound Description="laugh"/> <NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2583.439" EndTime="2585.890" Participant="me011">
it was advertised for performers and -
</Segment>
<Segment StartTime="2583.483" EndTime="2587.013" Participant="me013">
For the recor- for the record Adam is not a paid employee or a consultant of Crown.
</Segment>
<Segment StartTime="2586.130" EndTime="2588.386" Participant="me018">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2586.210" EndTime="2586.746" Participant="me011">
Excuse me?
</Segment>
<Segment StartTime="2586.380" EndTime="2586.959" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2587.013" EndTime="2588.085" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2587.431" EndTime="2588.069" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2587.785" EndTime="2589.039" Participant="fe008">
Oh. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2587.990" EndTime="2588.400" Participant="me011">
Excuse me?
</Segment>
<Segment StartTime="2588.069" EndTime="2588.746" Participant="fe016">
Right.
</Segment>
<Segment StartTime="2588.085" EndTime="2592.033" Participant="me013">
I said "For the record Adam is - is not a <Emphasis> paid </Emphasis> consultant or employee of Crown". <Comment Description="humor, mock legal disclaimer, specialized intonation"/>
</Segment>
<Segment StartTime="2589.065" EndTime="2589.463" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2589.475" EndTime="2589.865" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2589.914" EndTime="2591.070" Participant="me011">
That's right. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2590.420" EndTime="2594.134" Participant="fe016">
However, he may be solicited after these meetings are distributed.
</Segment>
<Segment StartTime="2591.688" EndTime="2594.225" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2591.947" EndTime="2594.565" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2592.033" EndTime="2594.516" Participant="me013">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2592.112" EndTime="2593.787" Participant="me018">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2592.360" EndTime="2593.653" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2594.285" EndTime="2595.080" Participant="me018">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2594.531" EndTime="2595.205" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2595.080" EndTime="2595.640" Participant="fe016">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2595.750" EndTime="2598.101" Participant="me011">
Well we're using the Crown P_Z_Ms. These are Crown aren't they?
</Segment>
<Segment StartTime="2596.410" EndTime="2598.682" Participant="fe016">
Don't worry about finishing your dissertation.
</Segment>
<Segment StartTime="2596.835" EndTime="2597.222" Participant="me013">
Yeah.
</Segment>
<Segment StartTime="2598.330" EndTime="2599.804" Participant="me013">
<Pronounce Pronunciation="&quot;white&quot;"> Right. </Pronounce>
</Segment>
<Segment StartTime="2598.420" EndTime="2599.795" Participant="me011">
The P_Z_Ms <Emphasis> are </Emphasis> <Emphasis> Crown, </Emphasis> aren't they?
</Segment>
<Segment StartTime="2599.804" EndTime="2600.401" Participant="me013">
<Emphasis> Yeah. </Emphasis>
</Segment>
<Segment StartTime="2600.041" EndTime="2602.346" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2600.372" EndTime="2601.289" Participant="me011">
Yeah, I thought they were.
</Segment>
<Segment StartTime="2600.981" EndTime="2602.141" Participant="me013">
You bet. You bet.
</Segment>
<Segment StartTime="2602.566" EndTime="2603.084" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2602.724" EndTime="2605.763" Participant="me011">
And they work <Emphasis> very </Emphasis> well. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2604.179" EndTime="2605.249" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2604.322" EndTime="2605.558" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2604.349" EndTime="2605.066" Participant="fe016">
Yes.
</Segment>
<Segment StartTime="2604.578" EndTime="2605.475" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2605.473" EndTime="2605.810" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2605.656" EndTime="2610.089" Participant="me013">
So if we go to a workshop about all this - this it's gonna be a meeting about meetings about meetings.
</Segment>
<Segment StartTime="2609.960" EndTime="2611.037" Participant="me018">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2609.973" EndTime="2611.188" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2610.011" EndTime="2612.529" Participant="fe008">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2610.094" EndTime="2610.731" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2610.146" EndTime="2610.864" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2610.292" EndTime="2611.548" Participant="fe016">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2612.007" EndTime="2613.220" Participant="me011">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2612.521" EndTime="2612.877" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2612.553" EndTime="2613.068" Participant="fe016">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2612.866" EndTime="2615.045" Participant="me013">
O_K. <VocalSound Description="laugh"/> So. <Comment Description="while laughing"/>
</Segment>
<Segment StartTime="2614.437" EndTime="2617.814" Participant="me011">
And then it - we have to go to the planning session for that workshop. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2616.871" EndTime="2619.939" Participant="me013">
Oh, yeah, what - Which'll be the meeting about the meeting about the meeting. Yeah?
</Segment>
<Segment StartTime="2616.955" EndTime="2618.788" Participant="me018">
Oh, god. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2617.136" EndTime="2618.889" Participant="fe016">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="2617.522" EndTime="2618.501" Participant="me001">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2618.332" EndTime="2621.450" Participant="me011">
Cuz then it would be a meeting about the meeting about the meeting about meetings.
</Segment>
<Segment StartTime="2619.061" EndTime="2619.616" Participant="fe008">
Ooo. <Comment Description="interjection"/>
</Segment>
<Segment StartTime="2620.502" EndTime="2623.092" Participant="me013">
Just start saying "M_ four". Yeah, O_K.
</Segment>
<Segment StartTime="2622.667" EndTime="2623.690" Participant="me011">
Yeah. <VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2622.713" EndTime="2623.550" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2622.832" EndTime="2623.456" Participant="me018">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2623.690" EndTime="2624.850" Participant="me011">
M_ to the fourth.
</Segment>
<Segment StartTime="2625.579" EndTime="2626.448" Participant="me013">
Should we do the digits?
</Segment>
<Segment StartTime="2626.457" EndTime="2627.215" Participant="me011">
Yep, go for it.
</Segment>
<Segment StartTime="2627.485" EndTime="2627.930" Participant="me013">
O_K.
</Segment>
<Segment StartTime="2628.750" EndTime="2632.372" Participant="me013" DigitTask="true">
Transcript <Pause/> two six one one, two six three zero. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2633.447" EndTime="2636.000" Participant="me013" DigitTask="true">
three two zero five six five three <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2636.457" EndTime="2638.127" Participant="me013" DigitTask="true">
four four five zero <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2638.425" EndTime="2640.446" Participant="me013" DigitTask="true">
five seven five six six <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2640.801" EndTime="2642.048" Participant="me013" DigitTask="true">
six six <Emphasis> two </Emphasis> <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2642.470" EndTime="2643.155" Participant="me013" DigitTask="true">
seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2643.663" EndTime="2644.102" Participant="me013" DigitTask="true">
eight <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2644.506" EndTime="2645.130" Participant="me013" DigitTask="true">
nine <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2645.543" EndTime="2646.590" Participant="me013" DigitTask="true">
zero two <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2646.800" EndTime="2648.438" Participant="me013" DigitTask="true">
one five two six three <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2648.750" EndTime="2650.148" Participant="me013" DigitTask="true">
two five one <Emphasis> two </Emphasis> <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2650.412" EndTime="2651.377" Participant="me013" DigitTask="true">
three seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2651.817" EndTime="2654.347" Participant="me013" DigitTask="true">
four seven zero six eight zero zero <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2654.769" EndTime="2655.578" Participant="me013" DigitTask="true">
five <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2655.806" EndTime="2656.579" Participant="me013" DigitTask="true">
six <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2656.984" EndTime="2657.968" Participant="me013" DigitTask="true">
eight one <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2657.968" EndTime="2658.917" Participant="me013" DigitTask="true">
nine three <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2659.356" EndTime="2659.971" Participant="me013" DigitTask="true">
O_ <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2660.340" EndTime="2661.702" Participant="me013" DigitTask="true">
zero four zero <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2661.702" EndTime="2662.493" Participant="me013" DigitTask="true">
one <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2662.493" EndTime="2663.108" Participant="me013" DigitTask="true">
two. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2664.396" EndTime="2667.853" Participant="me018" DigitTask="true">
Transcript two five nine one dash two six one zero. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2668.537" EndTime="2671.025" Participant="me018" DigitTask="true">
two four nine seven nine eight O_ <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2671.500" EndTime="2673.103" Participant="me018" DigitTask="true">
three four O_ six <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2672.024" EndTime="2674.375" Participant="mn014">
S- <Pause/> s-
</Segment>
<Segment StartTime="2673.499" EndTime="2674.520" Participant="me018" DigitTask="true">
four six <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2674.894" EndTime="2675.592" Participant="me018" DigitTask="true">
five <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2675.950" EndTime="2676.787" Participant="me018" DigitTask="true">
six <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2677.066" EndTime="2679.576" Participant="me018" DigitTask="true">
seven zero eight, seven six two one <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2679.975" EndTime="2680.851" Participant="me018" DigitTask="true">
nine three <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2681.584" EndTime="2682.832" Participant="me018" DigitTask="true">
O_ two O_ five <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2683.422" EndTime="2685.571" Participant="me018" DigitTask="true">
zero four one three one zero zero <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2686.389" EndTime="2687.807" Participant="me018" DigitTask="true">
one seven five four <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2688.421" EndTime="2690.553" Participant="me018" DigitTask="true">
two eight zero, two eight one four <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2691.210" EndTime="2691.771" Participant="me018" DigitTask="true">
three <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2692.386" EndTime="2693.594" Participant="me018" DigitTask="true">
four zero five <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2694.020" EndTime="2695.195" Participant="me018" DigitTask="true">
six two zero <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2695.785" EndTime="2697.508" Participant="me018" DigitTask="true">
seven three nine one three <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2698.041" EndTime="2699.273" Participant="me018" DigitTask="true">
eight four one <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2699.412" EndTime="2701.843" Participant="me018" DigitTask="true">
nine seven two, zero one zero zero <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2702.281" EndTime="2702.922" Participant="me018" DigitTask="true">
O_ <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2703.337" EndTime="2703.963" Participant="me018" DigitTask="true">
zero <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2704.342" EndTime="2706.563" Participant="me018" DigitTask="true">
one zero six seven nine three one <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2709.060" EndTime="2711.863" Participant="me011" DigitTask="true">
Transcript three seven three one, three seven five zero. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2712.622" EndTime="2715.013" Participant="me011" DigitTask="true">
seven two O_ O_, four nine nine <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2715.491" EndTime="2716.328" Participant="me011" DigitTask="true">
eight eight eight <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2716.842" EndTime="2717.440" Participant="me011" DigitTask="true">
<Emphasis> nine </Emphasis> eight <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2717.938" EndTime="2718.453" Participant="me011" DigitTask="true">
O_ <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2718.872" EndTime="2719.728" Participant="me011" DigitTask="true">
zero zero <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2720.074" EndTime="2721.441" Participant="me011" DigitTask="true">
two one one six <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2721.926" EndTime="2722.883" Participant="me011" DigitTask="true">
three three seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2723.560" EndTime="2724.377" Participant="me011" DigitTask="true">
<Emphasis> four </Emphasis> seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2724.855" EndTime="2726.424" Participant="me011" DigitTask="true">
five nine three O_ four <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2726.392" EndTime="2728.800" Participant="me013">
<VocalSound Description="breath"/>
</Segment>
<Segment StartTime="2726.931" EndTime="2728.333" Participant="me011" DigitTask="true">
six nine nine two two <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2728.851" EndTime="2729.521" Participant="me011" DigitTask="true">
seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2730.047" EndTime="2730.553" Participant="me011" DigitTask="true">
eight <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2730.972" EndTime="2731.749" Participant="me011" DigitTask="true">
O_ one <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2732.048" EndTime="2732.923" Participant="me011" DigitTask="true">
zero four <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2733.302" EndTime="2734.860" Participant="me011" DigitTask="true">
one nine five five seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2735.301" EndTime="2736.477" Participant="me011" DigitTask="true">
two nine nine O_ <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2736.930" EndTime="2737.410" Participant="me011" DigitTask="true">
three <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2737.725" EndTime="2738.347" Participant="me011" DigitTask="true">
four <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2738.851" EndTime="2739.948" Participant="me011" DigitTask="true">
five zero nine <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2740.352" EndTime="2742.793" Participant="me011" DigitTask="true">
seven one four one two zero zero. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2744.165" EndTime="2747.594" Participant="fe016" DigitTask="true">
Transcript three seven one one, three seven three O_. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2748.280" EndTime="2748.884" Participant="fe016" DigitTask="true">
six <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2748.884" EndTime="2750.418" Participant="fe016" DigitTask="true">
seven nine O_ three seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2750.558" EndTime="2751.056" Participant="fe016" DigitTask="true">
eight <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2751.056" EndTime="2751.654" Participant="fe016" DigitTask="true">
nine <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2751.654" EndTime="2753.666" Participant="fe016" DigitTask="true">
zero <Pause/> four five four <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2753.885" EndTime="2755.364" Participant="fe016" DigitTask="true">
one two four three nine <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2755.459" EndTime="2756.283" Participant="fe016" DigitTask="true">
two six <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2756.495" EndTime="2757.013" Participant="fe016" DigitTask="true">
three <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2757.133" EndTime="2757.671" Participant="fe016" DigitTask="true">
four <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2757.671" EndTime="2758.149" Participant="fe016" DigitTask="true">
five <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2758.149" EndTime="2759.304" Participant="fe016" DigitTask="true">
six O_ nine five <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2759.304" EndTime="2760.001" Participant="fe016" DigitTask="true">
eight one <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2760.321" EndTime="2762.683" Participant="fe016" DigitTask="true">
nine six six <Pause/> three eight one zero <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2763.261" EndTime="2764.218" Participant="fe016" DigitTask="true">
O_ five seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2764.218" EndTime="2764.745" Participant="fe016" DigitTask="true">
zero <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2764.745" EndTime="2765.164" Participant="fe016" DigitTask="true">
one <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2765.164" EndTime="2765.581" Participant="fe016" DigitTask="true">
two <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2765.581" EndTime="2767.508" Participant="fe016" DigitTask="true">
three <Pause/> zero seven zero eight <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2767.508" EndTime="2767.810" Participant="fe016">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2767.810" EndTime="2769.888" Participant="fe016" DigitTask="true">
five three one four eight O_ three <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2770.127" EndTime="2771.124" Participant="fe016" DigitTask="true">
six four two. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2771.582" EndTime="2773.128" Participant="me011">
Pause between the lines, remember? <Comment Description="whispered"/>
</Segment>
<Segment StartTime="2774.319" EndTime="2775.056" Participant="mn014">
<VocalSound Description="laugh"/>
</Segment>
<Segment StartTime="2774.855" EndTime="2777.290" Participant="fe008" DigitTask="true">
Two one seven one dash two one nine O_. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2776.967" EndTime="2777.567" Participant="mn014">
<VocalSound Description="breath-laugh"/>
</Segment>
<Segment StartTime="2777.927" EndTime="2779.304" Participant="fe008" DigitTask="true">
six two nine seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2780.075" EndTime="2782.238" Participant="fe008" DigitTask="true">
seven two seven O_ seven nine seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2782.856" EndTime="2783.860" Participant="fe008" DigitTask="true">
eight six <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2784.169" EndTime="2785.347" Participant="fe008" DigitTask="true">
nine nine O_ <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2786.004" EndTime="2787.109" Participant="fe008" DigitTask="true">
O_ eight nine <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2787.549" EndTime="2788.128" Participant="fe008" DigitTask="true">
zero <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2788.534" EndTime="2789.654" Participant="fe008" DigitTask="true">
one zero seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2789.963" EndTime="2791.237" Participant="fe008" DigitTask="true">
three two three one <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2791.643" EndTime="2792.876" Participant="fe008" DigitTask="true">
four two two four <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2793.370" EndTime="2794.996" Participant="fe008" DigitTask="true">
five seven seven eight four <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2795.451" EndTime="2797.458" Participant="fe008" DigitTask="true">
six eight zero seven <Pause/> zero <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2797.902" EndTime="2798.481" Participant="fe008" DigitTask="true">
seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2798.810" EndTime="2800.577" Participant="fe008" DigitTask="true">
eight O_ three four seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2800.809" EndTime="2801.427" Participant="fe008" DigitTask="true">
nine <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2801.833" EndTime="2802.975" Participant="fe008" DigitTask="true">
zero two zero <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2803.351" EndTime="2804.996" Participant="fe008" DigitTask="true">
one seven seven O_ seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2805.440" EndTime="2806.917" Participant="fe008" DigitTask="true">
two three O_ eight <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2807.338" EndTime="2808.231" Participant="fe008" DigitTask="true">
three nine <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2808.555" EndTime="2809.381" Participant="fe008" DigitTask="true">
<Emphasis> four </Emphasis> nine <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2809.752" EndTime="2810.385" Participant="fe008" DigitTask="true">
five. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2811.594" EndTime="2816.362" Participant="mn014" DigitTask="true">
Transcript two one five one dash two one seven O_ <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2817.079" EndTime="2818.874" Participant="mn014" DigitTask="true">
five two four eight <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2818.997" EndTime="2819.849" Participant="mn014" DigitTask="true">
six three <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2819.854" EndTime="2821.588" Participant="me001">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2820.272" EndTime="2823.241" Participant="mn014" DigitTask="true">
seven six eight seven, five six nine <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2823.241" EndTime="2823.938" Participant="mn014" DigitTask="true">
eight <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2824.476" EndTime="2825.215" Participant="mn014" DigitTask="true">
nine <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2825.651" EndTime="2827.190" Participant="mn014" DigitTask="true">
O_ O_ seven eight <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2827.584" EndTime="2828.326" Participant="mn014" DigitTask="true">
one one <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2828.899" EndTime="2830.330" Participant="mn014" DigitTask="true">
two three six <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2830.330" EndTime="2831.360" Participant="mn014" DigitTask="true">
three eight <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2831.827" EndTime="2834.167" Participant="mn014" DigitTask="true">
four eight one three O_ O_ two <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2834.637" EndTime="2835.334" Participant="mn014" DigitTask="true">
five <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2835.892" EndTime="2836.609" Participant="mn014" DigitTask="true">
six <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2837.127" EndTime="2837.772" Participant="mn014" DigitTask="true">
seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2838.283" EndTime="2840.454" Participant="mn014" DigitTask="true">
nine two eight, three five two eight <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2840.992" EndTime="2842.318" Participant="mn014" DigitTask="true">
O_ three O_ <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2842.726" EndTime="2843.782" Participant="mn014" DigitTask="true">
zero nine three <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2844.419" EndTime="2846.672" Participant="mn014" DigitTask="true">
one eight zero, eight two three two <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2846.991" EndTime="2847.509" Participant="mn014" DigitTask="true">
two <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2847.948" EndTime="2848.823" Participant="mn014" DigitTask="true">
three O_ <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2849.341" EndTime="2851.672" Participant="mn014" DigitTask="true">
four O_, six O_, nine six five <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2852.274" EndTime="2852.992" Participant="me001">
<VocalSound Description="clear throat"/>
</Segment>
<Segment StartTime="2853.868" EndTime="2856.717" Participant="me001" DigitTask="true">
Transcript two three five one dash two three seven zero. <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2857.385" EndTime="2857.893" Participant="me001" DigitTask="true">
two <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2858.271" EndTime="2858.897" Participant="me001" DigitTask="true">
three <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2859.678" EndTime="2861.288" Participant="me001" DigitTask="true">
five four zero four two <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2861.488" EndTime="2862.643" Participant="me001" DigitTask="true">
six four one <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2862.900" EndTime="2865.431" Participant="me001" DigitTask="true">
seven seven four eight seven <Emphasis> two </Emphasis> one <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2865.530" EndTime="2866.246" Participant="me001" DigitTask="true">
eight <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2866.407" EndTime="2867.090" Participant="me001" DigitTask="true">
nine <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2867.602" EndTime="2868.137" Participant="me001" DigitTask="true">
O_ <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2868.479" EndTime="2869.474" Participant="me001" DigitTask="true">
zero zero <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2869.853" EndTime="2870.754" Participant="me001" DigitTask="true">
two three <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2871.217" EndTime="2872.236" Participant="me001" DigitTask="true">
three three O_ <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2872.698" EndTime="2874.028" Participant="me001" DigitTask="true">
four eight two <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2874.456" EndTime="2875.539" Participant="me001" DigitTask="true">
five eight nine <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2876.060" EndTime="2878.196" Participant="me001" DigitTask="true">
six O_ nine on four O_ two <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2878.426" EndTime="2879.123" Participant="me001" DigitTask="true">
seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2879.593" EndTime="2881.983" Participant="me001" DigitTask="true">
eight zero nine eight zero four zero <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2882.571" EndTime="2883.724" Participant="me001" DigitTask="true">
O_ one seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2883.963" EndTime="2886.654" Participant="me001" DigitTask="true">
zero four three zero two eight one <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2886.663" EndTime="2886.931" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2887.098" EndTime="2889.298" Participant="me001" DigitTask="true">
one seven eight five seven <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2889.613" EndTime="2891.673" Participant="me001" DigitTask="true">
two nine six two O_ - <Comment Description="corrected below"/> <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2892.774" EndTime="2893.215" Participant="me001">
Excuse me.
</Segment>
<Segment StartTime="2893.215" EndTime="2895.583" Participant="me001" DigitTask="true">
two nine six two <Emphasis> one </Emphasis> one O_ <Comment Description="Digits"/>
</Segment>
<Segment StartTime="2895.386" EndTime="2899.311" Participant="me011">
<NonVocalSound Description="mike noise"/> O_K. <NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2897.315" EndTime="2898.123" Participant="me013">
<Pronounce Pronunciation="&quot;mm-kay&quot;"> O_K. </Pronounce>
</Segment>
<Segment StartTime="2898.488" EndTime="2898.866" Participant="fe016">
Huh.
</Segment>
<Segment StartTime="2898.866" EndTime="2902.220" Participant="fe016">
<NonVocalSound Description="mike noise"/>
</Segment>
<Segment StartTime="2899.391" EndTime="2899.886" Participant="me001">
<VocalSound Description="clear throat"/>
</Segment>
<Segment StartTime="2900.380" EndTime="2906.436" Participant="fe008">
<NonVocalSound Description="mike noise"/>
</Segment>
</Transcript>
</Meeting>