Datasets:
image unknown | image_format stringclasses 1
value | source stringclasses 1
value | markered bool 1
class | capture stringclasses 7
values | split stringclasses 1
value | height int32 320 320 | width int32 240 240 | obj_name stringclasses 1
value | init_pose int32 | side stringclasses 0
values | x_mm float32 | y_mm float32 | z_mm float32 | quat_x float32 | quat_y float32 | quat_z float32 | quat_w float32 | indenter stringclasses 1
value | indenter_param stringclasses 0
values | f_x float32 | f_y float32 | f_z float32 | grid_z_max float32 | grid_z_mean float32 | episode stringclasses 7
values | frame_idx int32 0 5k | digit_class int32 | gel_variant stringclasses 1
value | domain stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
[
255,
216,
255,
224,
0,
16,
74,
70,
73,
70,
0,
1,
1,
0,
0,
1,
0,
1,
0,
0,
255,
219,
0,
67,
0,
3,
2,
2,
2,
2,
2,
3,
2,
2,
2,
3,
3,
3,
3,
4,
6,
4,
4,
4,
4,
4,
8,
6,
6,
5,
6,
9,
8,
10,
10,
9,
8,
9,
9,
1... | jpeg | sparsh | false | flat_batch_1_f00 | flat | 320 | 240 | indenter_flat | null | null | null | null | null | null | null | null | null | flat | null | null | null | null | null | null | batch_1_f00 | 20 | null | markerless | real |
[
255,
216,
255,
224,
0,
16,
74,
70,
73,
70,
0,
1,
1,
0,
0,
1,
0,
1,
0,
0,
255,
219,
0,
67,
0,
3,
2,
2,
2,
2,
2,
3,
2,
2,
2,
3,
3,
3,
3,
4,
6,
4,
4,
4,
4,
4,
8,
6,
6,
5,
6,
9,
8,
10,
10,
9,
8,
9,
9,
1... | jpeg | sparsh | false | flat_batch_1_f00 | flat | 320 | 240 | indenter_flat | null | null | null | null | null | null | null | null | null | flat | null | null | null | null | null | null | batch_1_f00 | 35 | null | markerless | real |
[
255,
216,
255,
224,
0,
16,
74,
70,
73,
70,
0,
1,
1,
0,
0,
1,
0,
1,
0,
0,
255,
219,
0,
67,
0,
3,
2,
2,
2,
2,
2,
3,
2,
2,
2,
3,
3,
3,
3,
4,
6,
4,
4,
4,
4,
4,
8,
6,
6,
5,
6,
9,
8,
10,
10,
9,
8,
9,
9,
1... | jpeg | sparsh | false | flat_batch_1_f00 | flat | 320 | 240 | indenter_flat | null | null | null | null | null | null | null | null | null | flat | null | null | null | null | null | null | batch_1_f00 | 36 | null | markerless | real |
[
255,
216,
255,
224,
0,
16,
74,
70,
73,
70,
0,
1,
1,
0,
0,
1,
0,
1,
0,
0,
255,
219,
0,
67,
0,
3,
2,
2,
2,
2,
2,
3,
2,
2,
2,
3,
3,
3,
3,
4,
6,
4,
4,
4,
4,
4,
8,
6,
6,
5,
6,
9,
8,
10,
10,
9,
8,
9,
9,
1... | jpeg | sparsh | false | flat_batch_1_f00 | flat | 320 | 240 | indenter_flat | null | null | null | null | null | null | null | null | null | flat | null | null | null | null | null | null | batch_1_f00 | 37 | null | markerless | real |
[
255,
216,
255,
224,
0,
16,
74,
70,
73,
70,
0,
1,
1,
0,
0,
1,
0,
1,
0,
0,
255,
219,
0,
67,
0,
3,
2,
2,
2,
2,
2,
3,
2,
2,
2,
3,
3,
3,
3,
4,
6,
4,
4,
4,
4,
4,
8,
6,
6,
5,
6,
9,
8,
10,
10,
9,
8,
9,
9,
1... | jpeg | sparsh | false | flat_batch_1_f00 | flat | 320 | 240 | indenter_flat | null | null | null | null | null | null | null | null | null | flat | null | null | null | null | null | null | batch_1_f00 | 38 | null | markerless | real |
[
255,
216,
255,
224,
0,
16,
74,
70,
73,
70,
0,
1,
1,
0,
0,
1,
0,
1,
0,
0,
255,
219,
0,
67,
0,
3,
2,
2,
2,
2,
2,
3,
2,
2,
2,
3,
3,
3,
3,
4,
6,
4,
4,
4,
4,
4,
8,
6,
6,
5,
6,
9,
8,
10,
10,
9,
8,
9,
9,
1... | jpeg | sparsh | false | flat_batch_1_f00 | flat | 320 | 240 | indenter_flat | null | null | null | null | null | null | null | null | null | flat | null | null | null | null | null | null | batch_1_f00 | 39 | null | markerless | real |
[
255,
216,
255,
224,
0,
16,
74,
70,
73,
70,
0,
1,
1,
0,
0,
1,
0,
1,
0,
0,
255,
219,
0,
67,
0,
3,
2,
2,
2,
2,
2,
3,
2,
2,
2,
3,
3,
3,
3,
4,
6,
4,
4,
4,
4,
4,
8,
6,
6,
5,
6,
9,
8,
10,
10,
9,
8,
9,
9,
1... | jpeg | sparsh | false | flat_batch_1_f00 | flat | 320 | 240 | indenter_flat | null | null | null | null | null | null | null | null | null | flat | null | null | null | null | null | null | batch_1_f00 | 40 | null | markerless | real |
[
255,
216,
255,
224,
0,
16,
74,
70,
73,
70,
0,
1,
1,
0,
0,
1,
0,
1,
0,
0,
255,
219,
0,
67,
0,
3,
2,
2,
2,
2,
2,
3,
2,
2,
2,
3,
3,
3,
3,
4,
6,
4,
4,
4,
4,
4,
8,
6,
6,
5,
6,
9,
8,
10,
10,
9,
8,
9,
9,
1... | jpeg | sparsh | false | flat_batch_1_f00 | flat | 320 | 240 | indenter_flat | null | null | null | null | null | null | null | null | null | flat | null | null | null | null | null | null | batch_1_f00 | 41 | null | markerless | real |
[
255,
216,
255,
224,
0,
16,
74,
70,
73,
70,
0,
1,
1,
0,
0,
1,
0,
1,
0,
0,
255,
219,
0,
67,
0,
3,
2,
2,
2,
2,
2,
3,
2,
2,
2,
3,
3,
3,
3,
4,
6,
4,
4,
4,
4,
4,
8,
6,
6,
5,
6,
9,
8,
10,
10,
9,
8,
9,
9,
1... | jpeg | sparsh | false | flat_batch_1_f00 | flat | 320 | 240 | indenter_flat | null | null | null | null | null | null | null | null | null | flat | null | null | null | null | null | null | batch_1_f00 | 42 | null | markerless | real |
[
255,
216,
255,
224,
0,
16,
74,
70,
73,
70,
0,
1,
1,
0,
0,
1,
0,
1,
0,
0,
255,
219,
0,
67,
0,
3,
2,
2,
2,
2,
2,
3,
2,
2,
2,
3,
3,
3,
3,
4,
6,
4,
4,
4,
4,
4,
8,
6,
6,
5,
6,
9,
8,
10,
10,
9,
8,
9,
9,
1... | jpeg | sparsh | false | flat_batch_1_f00 | flat | 320 | 240 | indenter_flat | null | null | null | null | null | null | null | null | null | flat | null | null | null | null | null | null | batch_1_f00 | 43 | null | markerless | real |
[
255,
216,
255,
224,
0,
16,
74,
70,
73,
70,
0,
1,
1,
0,
0,
1,
0,
1,
0,
0,
255,
219,
0,
67,
0,
3,
2,
2,
2,
2,
2,
3,
2,
2,
2,
3,
3,
3,
3,
4,
6,
4,
4,
4,
4,
4,
8,
6,
6,
5,
6,
9,
8,
10,
10,
9,
8,
9,
9,
1... | jpeg | sparsh | false | flat_batch_1_f00 | flat | 320 | 240 | indenter_flat | null | null | null | null | null | null | null | null | null | flat | null | null | null | null | null | null | batch_1_f00 | 44 | null | markerless | real |
[
255,
216,
255,
224,
0,
16,
74,
70,
73,
70,
0,
1,
1,
0,
0,
1,
0,
1,
0,
0,
255,
219,
0,
67,
0,
3,
2,
2,
2,
2,
2,
3,
2,
2,
2,
3,
3,
3,
3,
4,
6,
4,
4,
4,
4,
4,
8,
6,
6,
5,
6,
9,
8,
10,
10,
9,
8,
9,
9,
1... | jpeg | sparsh | false | flat_batch_1_f00 | flat | 320 | 240 | indenter_flat | null | null | null | null | null | null | null | null | null | flat | null | null | null | null | null | null | batch_1_f00 | 45 | null | markerless | real |
[
255,
216,
255,
224,
0,
16,
74,
70,
73,
70,
0,
1,
1,
0,
0,
1,
0,
1,
0,
0,
255,
219,
0,
67,
0,
3,
2,
2,
2,
2,
2,
3,
2,
2,
2,
3,
3,
3,
3,
4,
6,
4,
4,
4,
4,
4,
8,
6,
6,
5,
6,
9,
8,
10,
10,
9,
8,
9,
9,
1... | jpeg | sparsh | false | flat_batch_1_f00 | flat | 320 | 240 | indenter_flat | null | null | null | null | null | null | null | null | null | flat | null | null | null | null | null | null | batch_1_f00 | 46 | null | markerless | real |
End of preview. Expand in Data Studio
GelSight Mini Pretrain · Non-Commercial Extension
⚠️ Non-commercial use only. This repository is licensed CC-BY-NC-4.0 because it includes upstream sources whose licenses restrict commercial use. For commercial-friendly Mini tactile data, see the main yxma/gelsight-mini-pretrain repo (CC-BY-4.0).
This dataset is the CC-BY-NC extension of yxma/gelsight-mini-pretrain.
It contains only the GelSight Mini sources whose upstream licenses are
not compatible with CC-BY-4.0 aggregation. All data is processed
through the same pipeline as the main repo (same schema, same area+intensity
validity filter, same channel-order normalization), so users can simply
load both repos and concatenate them to get a larger pool.
Sample images
Sparsh (3 indenter shapes)
| Indenter | Frames | Random samples |
|---|---|---|
| flat | 8,056 |  |
| sharp | 11,844 |  |
| sphere | 46,544 |  |
When to use this repo
Use the main repo if:
- You want commercial use rights
- You only need ~830K frames across 12 sources (already plenty for VAE pretraining)
Use both (main + this extension) if:
- Your work is non-commercial (academic, internal research)
- You want maximum coverage of public Mini tactile data
- You're OK respecting the NC clause on the extension subsets only
How to combine with the main repo
from datasets import concatenate_datasets, load_dataset
# Load all subsets from the main repo (CC-BY-4.0)
main_real = concatenate_datasets([
load_dataset("yxma/gelsight-mini-pretrain", c, split="train")
for c in ["fota_labeled", "fota_unlabeled", "threedcal", "feats",
"gelslam", "tactile_tracking", "real_tactile_mnist",
"feelanyforce", "unit", "tacquad"]
])
# Add NC extension (sparsh has 3 indenter-named splits)
nc_extras = concatenate_datasets([
load_dataset("yxma/gelsight-mini-pretrain-nc", "sparsh", split=s)
for s in ["flat", "sharp", "sphere"]
])
all_markerless = concatenate_datasets([main_real, nc_extras])
print(f"Total: {len(all_markerless):,} frames")
Schema
Identical to the main repo (30 columns, parquet, JPEG q=92 images).
Every row has domain="real", markered=False,
gel_variant="markerless", plus source-specific metadata.
Pipeline parity
Applies the same unified area+intensity validity filter as the main repo:
pixel_diff = |frame - baseline|on central 50% crop, greyscalemask = pixel_diff > 10(sensor noise floor)area = mask.sum(),intensity = pixel_diff[mask].mean()- KEEP iff
area >= 40 px AND intensity >= 12 grey-levels - ELSE keep with probability
1.5%(background diversity)
Channel-order normalization
Sparsh's upstream pickles contain a mix of RGB and BGR images
(apparently from heterogeneous data-collection pipelines at Meta). We
normalize all images to RGB by checking each image's per-channel means
and swapping R<->B when mean(R) > mean(B) (GelSight Mini's at-rest
gel is illuminated such that B > R is the correct channel order).
After normalization, every image in this repo is guaranteed RGB.
Sources currently included
| Subset | Upstream | License | Frames | Notes |
|---|---|---|---|---|
sparsh |
facebook/SparshGelSight |
CC-BY-NC-4.0 | 66,444 (flat 8K + sharp 12K + sphere 47K) | Part of Meta's Sparsh / TacBench. GelSight Mini, markerless, 3 indenter shapes with paired ATI nano17 force ground truth. Includes channel-order normalization (upstream pkls mix RGB and BGR). |
Deprecated: The earlier
faf_force_estimation/subset (fromfacebook/gelsight-force-estimation) has been removed -- it was a smaller snapshot of the same collection protocol and data assparsh, just released under a different upstream repo name. Sparsh is the canonical larger version of the same protocol.
Citation chain
If you use this extension, please cite both the main aggregation and the upstream NC source:
@dataset{gelsight_mini_pretrain_nc,
title = {GelSight Mini Pretrain - NC extension},
author = {Ma, Yuxiang},
year = {2026},
url = {https://huggingface.co/datasets/yxma/gelsight-mini-pretrain-nc},
}
@misc{sparsh_2024,
title = {Sparsh: Self-supervised touch representations for vision-based tactile sensing},
author = {Higuera, Carolina and Sharma, Akash and Bodduluri, Chaithanya Krishna and Fan, Taosha and Lancaster, Patrick and Kalakrishnan, Mrinal and Kaess, Michael and Boots, Byron and Lambeta, Mike and Wu, Tingfan and Mukadam, Mustafa},
year = {2024},
url = {https://huggingface.co/datasets/facebook/SparshGelSight},
}
License
This aggregated release is CC-BY-NC-4.0 because it inherits the strictest license among the included sources. Component datasets retain their own original licenses; cite each upstream source individually.
- Downloads last month
- 45