---
pretty_name: DexCanvas
license: odbl
language:
- en
task_categories:
- robotics
tags:
- robotics
- manipulation
- dexterous-manipulation
- hand-object-interaction
- trajectory-data
- contact-forces
- physics-simulation
annotations_creators:
- machine-generated
size_categories:
- 100K<n<1M
viewer: false
configs:
- config_name: default
  default: true
  data_files:
  - split: train
    path: data/train-*.parquet
- config_name: image_captures
  data_files:
  - split: train
    path: metadata/image_captures.parquet
- config_name: trajectory_image_links
  data_files:
  - split: train
    path: metadata/trajectory_image_links.parquet
---

# DexCanvas

**DexCanvas: Bridging Human Demonstrations and Robot Learning for Dexterous Manipulation**

[Paper (arXiv:2510.15786)](https://arxiv.org/abs/2510.15786)

![DexCanvas cover](docs/cover.png)

DexCanvas is a dataset of **111,857 complete right-hand trajectories** and **55,442,683 frames**
of rigid hand–object interaction. Every Parquet row is one trajectory with a shared time axis for
hand kinematics, object pose, and per-frame contacts. For each trajectory, the repository also
provides the complete dual-camera JPEG stream of the corresponding capture and shape-specific
kinematic assets for replay.

Camera images are associated at the complete Raw trajectory level. The release does not claim
that trajectory frame `i` corresponds to image `i`.

## Summary

- 111,857 trajectories; 55,442,683 frames; 96,484,004 contact records.
- 32 `action_label` values spanning Cutkosky-style manipulation primitives, including palmar pinch,
  prismatic and lateral grasps, sphere grasps, extension, adduction, tripod, writing tripod,
  ring, distal type, and index-finger extension.
- 22 object categories spanning geometric primitives and household items: banana, bowl, cube1,
  cube2, cuboid1, cuboid2, cuboid3, cylinder1, cylinder2, cylinder3, cylinder4, cylinder5,
  cylinder6, cylinder7, iphone, largeclamp, mayonnaisebottle, pitcherbase, powerdrill, scissor,
  sphere1, sphere2.
- 21,079 deduplicated dual-camera captures; 22,212,754 JPEG frames.
- Eight right-hand MANO shape identities, each packaged as a 28-DoF floating-base URDF bundle;
  one canonical metre-scale mesh per object category.
- The package contains trajectory Parquet shards, image archives, capture/link metadata,
  kinematic assets, documentation, checksums, and a local Rerun replay tool.

## What is included

- **Trajectory data:** `data/train-*.parquet`, with `data/manifest.json` recording shard coverage,
  byte sizes, ID ranges, and SHA-256 values.
- **Capture images:** `images/<prefix>/<capture>.tar`, WebDataset-compatible archives containing
  the complete original dual-camera JPEG streams and adjacent public JSON sidecars.
- **Indexes:** `metadata/image_captures.parquet` and
  `metadata/trajectory_image_links.parquet` describe captures and trajectory-to-capture links.
- **Kinematic assets:** `assets/hands/hand_shape_*/mano_hand.urdf` with referenced meshes;
  `assets/objects/<category>/mesh.stl`; see `assets/manifest.json`.
- **Replay:** `replay/` contains a standalone Rerun player that reconstructs URDF link geometry
  through forward kinematics.
- **Integrity:** `SHA256SUMS`, `all_rows_report.json`, `image_validation_report.json`,
  `clean_room_report.json`, and `public_string_audit.json`.

## Use the data

Streaming is recommended because a row contains a complete trajectory:

```python
from datasets import load_dataset

trajectories = load_dataset(
    "Dex-GEM/DexCanvas",
    split="train",
    streaming=True,
)

trajectory = next(iter(trajectories))
print(trajectory["trajectory_id"], trajectory["num_frames"])
```

The Hub Dataset Viewer is disabled because complete nested trajectories can exceed its
per-row display limit.

## Row schema

```text
trajectory_id
action_label
frame_rate_hz
num_frames
timestamps_s[T]

hands[]
  ├── hand_id
  ├── side
  ├── mano_shape[10]
  ├── global_position_world_m[T, 3]
  ├── global_rotation_world_rotvec_rad[T, 3]
  ├── mano_pose_rotvec_rad[T, 48]
  ├── joint_positions_world_m[T, 21, 3]
  ├── urdf_dof[T, 28]
  └── urdf_dof_target[T, 28]

objects[]
  ├── object_id
  ├── category
  ├── position_world_m[T, 3]
  └── rotation_world_rotvec_rad[T, 3]

contacts[T][]
  ├── hand_id
  ├── joint_name
  ├── object_id
  ├── total_force_world_n[3]
  ├── total_force_wrist_n[3] or null
  ├── total_force_joint_n[3]
  ├── total_force_object_n[3]
  └── points[]
```

`timestamps_s` is authoritative. `frame_rate_hz` is calculated from its median interval rather
than copied from source metadata. Positions are in metres, axis-angle rotation vectors are in
radians, and forces are in newtons.

See [`DATA_SCHEMA.md`](DATA_SCHEMA.md) for the trajectory contract and
[`IMAGE_DATA.md`](IMAGE_DATA.md) for capture indexes, WebDataset archives, and the alignment
boundary.

## Assets

The asset manifest is [`assets/manifest.json`](assets/manifest.json).

- `assets/hands/hand_shape_*/mano_hand.urdf`: eight public shape IDs covering every hand shape
  in the data;
- `assets/hands/hand_shape_*/meshes/`: all visual and collision meshes referenced by each URDF;
- `assets/objects/<category>/mesh.stl`: one canonical metre-scale object mesh per category.

Public shape IDs are matched to the `mano_shape[10]` stored in each trajectory. No operator
identity is included.

## Rerun replay

Install the replay dependencies:

```bash
python -m venv .venv
. .venv/bin/activate
pip install -r replay/requirements.txt
```

Replay by public trajectory ID:

```bash
python replay/replay_rerun.py \
  --data data \
  --assets assets \
  --trajectory-id traj_00000001 \
  --spawn
```

Export a portable recording:

```bash
python replay/replay_rerun.py \
  --data data \
  --assets assets \
  --trajectory-id traj_00000001 \
  --save trajectory.rrd
```

The player reconstructs hand-link geometry through URDF forward kinematics and displays the
object mesh, URDF and recorded hand joints, contact points, and scaled contact-force arrows.
It uses only this repository's Parquet data and relative assets; Lance and private storage are
not required.

<p align="center">
  <img src="docs/replay_preview.gif" alt="Trajectory replay" width="480">
  <br>
  <em>Example: one trajectory replayed via <code>replay/replay_rerun.py</code>. Hand (skin), object (yellow), and contact points (red) are reconstructed entirely from Parquet data and package assets.</em>
</p>

## Dataset structure and integrity

- Split: `train`
- Trajectories: 111,857
- Frames: 55,442,683
- Contact records: 96,484,004
- Active hands per trajectory: 1
- Objects per trajectory: 1
- Unique hand shapes: 8
- Object categories: 22
- Deduplicated camera captures: 21,079 (two cameras each)
- Camera images: 22,212,754 JPEGs (approximately 889 GB of TAR archives)

Exact shard counts, byte sizes, ID ranges, and SHA-256 values are recorded in
[`data/manifest.json`](data/manifest.json). Package checksums are in [`SHA256SUMS`](SHA256SUMS).
Every published row is covered by the aggregate structural scan in
[`all_rows_report.json`](all_rows_report.json), image integrity is recorded in
[`image_validation_report.json`](image_validation_report.json), source-independent replay is
recorded in [`clean_room_report.json`](clean_room_report.json), and every public string field
is covered by [`public_string_audit.json`](public_string_audit.json).

## License

This release is made available under the [Open Data Commons Open Database License (ODbL) v1.0](https://opendatacommons.org/licenses/odbl/1-0/).
See [`NOTICE_ASSETS.md`](NOTICE_ASSETS.md) for third-party attribution covering the hand and
object assets.

## Citation

If you use DexCanvas, please cite:

```bibtex
@misc{xu2025dexcanvas,
      title={DexCanvas: Bridging Human Demonstrations and Robot Learning for Dexterous Manipulation},
      author={Xinyue Xu and Jieqiang Sun and Jing Dai and Siyuan Chen and Lanjie Ma and Ke Sun and Bin Zhao and Jianbo Yuan and Sheng Yi and Haohua Zhu and Irvine Lu},
      year={2025},
      eprint={2510.15786},
      archivePrefix={arXiv},
      primaryClass={cs.RO},
      url={https://arxiv.org/abs/2510.15786},
}
```