remove touchandgo: uses ORIGINAL GelSight (white illumination, 640x480 flat-channel R~G~B~137) not Mini. Verified via channel-mean signature mismatch with all other Mini sources (B>G>R with 30+ grey-level spread). Document reason in SOURCES.md 'Investigated but not included'. Bumps real total from 748,024 to 745,958 frames.

README.md

@@ -88,10 +88,6 @@ configs:
         path: tacquad/data_outdoor-*.parquet
       - split: data_fine
         path: tacquad/data_fine-*.parquet
-  - config_name: touchandgo
-    data_files:
-      - split: train
-        path: touchandgo/train-*.parquet
   - config_name: tvl
     data_files:
       - split: train
@@ -272,7 +268,6 @@ per-axis distributions.
 | `sim_starstruck`  | **SIM** · Taxim-rendered Mini imagery of star objects | **166,104** (150K train + 16K test; raw upstream + stride=3 + I_min=10) | markerless | episode |
 | `unit`            | UniT continuous 3D-pose tracking      | **11,340** | markerless | 3D-pose target (x,y,z,yaw) |
 | `tacquad`         | TacQuad quad-sensor benchmark (Mini stream) | **12,195** (5K indoor + 4K outdoor + 3K fine, 181 objects) | markerless | object name + environment |
-| `touchandgo`      | Touch-and-Go in-the-wild paired touch+RGB | **2,066** (8 sessions ingested; gdown rate-limited at ~9/many) | markerless | session id + paired RGB scene (RGB not stored) |
 | `tvl`             | TVL Touch-Vision-Language paired tactile+RGB+caption | **209,795** (HCT + SSVTP subsets; multi-resolution: 640×480 + 320×240) | markerless | session id + paired RGB + GPT-4V caption (RGB+text not stored) |
 
 ¹ FoTA used **different gels on the two gripper fingers** for many of its captures. Approximately 36 of 124 captures use a markered gel on the right finger and a markerless gel on the left; the remaining 88 captures use markerless gels on both. The per-row `markered` column was set by averaging ~50 frames per capture and counting visible dark dots in the mean image (threshold ≥10 dots). Use it to filter:

SOURCES.md

@@ -591,46 +591,7 @@ real-world objects) in the entire dataset.
 
 ---
 
-## 12 · Touch-and-Go — in-the-wild paired vision-touch (`touchandgo`)
-
-**Intro.** Touch-and-Go (Yang et al., NeurIPS 2022) is a multimodal
-dataset of human data collectors probing objects in natural environments
-with a GelSight Mini tactile sensor while recording egocentric RGB
-video. Each session has `gelsight.mp4` (tactile) + `video.mp4` (RGB
-scene). We ingest **only the tactile video stream**, decoding gelsight
-frames at stride-5 and applying the contact filter.
-
-**Source release.**
-- 🔗 [GitHub project page](https://github.com/fredfyyang/Touch-and-Go)
-- 🔗 Hosted on Google Drive (~5 GB across 250+ sessions, rate-limited)
-- 📜 License: CC-BY-4.0
-
-**Original format.** One folder per session under
-`touch_and_go/dataset/<session>/`, with `gelsight.mp4`, `video.mp4`,
-`time1.npy`, `time2.npy`.
-
-**How we processed it.** Decoded `gelsight.mp4` with imageio+ffmpeg at
-stride-5 (every 5th frame), applied unified contact filter (I_min=12,
-A_min=40, 1.5% bg), channel-order auto-normalized. Initial batch was
-limited by Google Drive's anti-abuse rate-limiting: gdown succeeded on
-8 / 250+ sessions before hitting the per-folder access cap. We have a
-retry loop running in the background that will incrementally pull more
-sessions; the parquet on HF will be re-pushed as additional sessions
-become available.
-
-**Stats after processing (initial batch).**
-
-| Subset       | Frames | Resolution | Markered / Markerless | Splits |
-|--------------|-------:|------------|----------------------:|--------|
-| `touchandgo` | 2,066  | 240 × 320  | 0 / 2,066             | train 2,066 |
-
-**40 random samples:**
-
-![touchandgo](assets/samples_40_touchandgo.png)
-
----
-
-## 13 · TVL — Touch-Vision-Language paired tactile + RGB + caption (`tvl`)
+## 12 · TVL — Touch-Vision-Language paired tactile + RGB + caption (`tvl`)
 
 **Intro.** TVL (Fu et al., ICML 2024) is a multimodal alignment dataset
 with ~44K paired vision-touch examples plus English captions (10%
@@ -686,8 +647,15 @@ smaller SSVTP contribution (~10K@240×320).
   [`yxma/gelsight-mini-pretrain-nc`](https://huggingface.co/datasets/yxma/gelsight-mini-pretrain-nc),
   where it is now superseded by `sparsh` — the same data at a larger
   upstream snapshot.)
-- **TVL — Touch-Vision-Language** (Yang et al. 2024) — has paired RGB +
-  caption labels. Not yet ingested. CC-BY-4.0; ~44K Mini frames available.
+- **Touch-and-Go** (Yang et al., NeurIPS 2022, [GitHub](https://github.com/fredfyyang/Touch-and-Go)) —
+  **uses a different GelSight model, not the Mini.** The dataset's
+  `gelsight.mp4` clips show 640×480 frames with white-light illumination
+  (channel means R/G/B ≈ 137/135/140, flat ~grey) — completely
+  different from the Mini's 3-colored-LED signature (Mini at-rest has
+  B > G > R with ~30 grey-level spread). We initially ingested 2,066
+  frames before noticing the channel-signature mismatch and verifying
+  against the paper. Removed from this aggregate; could go to a
+  separate "gelsight-large-pretrain" repo if needed.
 - **Touch and Go** (Yang et al. 2022) — has paired natural-scene RGB.
   Not yet ingested. CC-BY-4.0; ~13K Mini frames available.
 - **YCB-Sight** — CC-BY-SA-4.0 (viral copyleft), and the sim is not
@@ -737,11 +705,10 @@ B > R at rest, matching the Mini's reference illumination geometry.
 | `tacquad`             | real   | 12,195  | 0.11       | 320 × 240  | 0        | 12,195     |
 | `sim_tactile_mnist`   | sim    | 150,601 | 1.28       | 320 × 240  | 0        | 150,601    |
 | `sim_starstruck`      | sim    | 166,104 | 1.39       | 320 × 240  | 0        | 166,104    |
-| `touchandgo`          | real   |   2,066 | 0.07       | 240 × 320  | 0        |   2,066    |
 | `tvl`                 | real   | 209,795 | 7.90       | mixed¹     | 0        | 209,795    |
-| **Real total**        |        | **748,024** | **~13.4** |        | **63,586** | **684,438** |
+| **Real total**        |        | **745,958** | **~13.3** |        | **63,586** | **682,372** |
 | **Sim total**         |        | **316,705** | **~2.67** |        | **0**     | **316,705** |
-| **Grand total**       |        | **1,064,729** | **~16.1** |       | **63,586** | **1,001,143** |
+| **Grand total**       |        | **1,062,663** | **~16.0** |       | **63,586** | **999,077** |
 
 ![summary pies](assets/summary_pies.png)
 

touchandgo/train-00000-of-00001.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:cba983962da92630e45949181409ce16ffa40f5e9bc47236c78850099b07e1fc
-size 74096866