chenwangj commited on
Commit
a2b8ae1
Β·
verified Β·
1 Parent(s): a1abbb1

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +50 -0
README.md CHANGED
@@ -1,3 +1,53 @@
1
  ---
2
  license: cc-by-4.0
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: cc-by-4.0
3
  ---
4
+ # Dataset Card for DexCap-Data
5
+
6
+ ## Dataset Description
7
+
8
+ - **Homepage:** https://dex-cap.github.io/
9
+ - **Repository:** https://github.com/j96w/DexCap
10
+ - **Paper:** https://arxiv.org/abs/2403.07788
11
+
12
+ ### Dataset Summary
13
+
14
+ This is the official dataset collected by the DexCap system to train dexterous robot manipulation using human hand motion capture data, as presented in the [paper](https://arxiv.org/abs/2403.07788). It contains 30 minutes of mocap data for the wiping task and 60 minutes of in-the-wild mocap data for the packaging task.
15
+
16
+ ## Dataset Structure
17
+
18
+ Both raw data (`.zip`) and postprocessed data (`.hdf5`) are provided. The raw data is structured as follows:
19
+ ```
20
+ save_data_scenario_1
21
+ β”œβ”€β”€ frame_0
22
+ β”‚ β”œβ”€β”€ color_image.jpg # Chest camera RGB image
23
+ β”‚ β”œβ”€β”€ depth_image.png # Chest camera depth image
24
+ β”‚ β”œβ”€β”€ pose.txt # Chest camera 6-DoF pose in world frame
25
+ β”‚ β”œβ”€β”€ pose_2.txt # Left hand 6-DoF pose in world frame
26
+ β”‚ β”œβ”€β”€ pose_3.txt # Right hand 6_DoF pose in world frame
27
+ β”‚ β”œβ”€β”€ left_hand_joint.txt # Left hand joint positions (3D) in the palm frame
28
+ β”‚ └── right_hand_joint.txt # Right hand joint positions (3D) in the palm frame
29
+ β”œβ”€β”€ frame_1
30
+ └── ...
31
+ ```
32
+ Details on how to postprocess the raw data into a training dataset (`.hdf5`) are introduced in the [GitHub repo](https://github.com/j96w/DexCap). The training scripts for policy learning with the HDF5 dataset are also included in the repo.
33
+
34
+ ## Dataset Creation
35
+
36
+ All data are collected by the [DexCap system](https://dex-cap.github.io/).
37
+
38
+ ## Additional Information
39
+
40
+ ### Licensing Information
41
+
42
+ This dataset is released under the [Creative Commons Attribution 4.0 International (CC BY 4.0)](https://creativecommons.org/licenses/by/4.0/legalcode) license.
43
+
44
+ ### Citation Information
45
+
46
+ ```bibtex
47
+ @article{wang2024dexcap,
48
+ title = {DexCap: Scalable and Portable Mocap Data Collection System for Dexterous Manipulation},
49
+ author = {Wang, Chen and Shi, Haochen and Wang, Weizhuo and Zhang, Ruohan and Fei-Fei, Li and Liu, C. Karen},
50
+ journal = {arXiv preprint arXiv:2403.07788},
51
+ year = {2024}
52
+ }
53
+ ```