The dataset viewer is not available for this split.
Error code: JobManagerCrashedError
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
MoCapAct Dataset
Control of simulated humanoid characters is a challenging benchmark for sequential decision-making methods, as it assesses a policyβs ability to drive an inherently unstable, discontinuous, and high-dimensional physical system. Motion capture (MoCap) data can be very helpful in learning sophisticated locomotion policies by teaching a humanoid agent low-level skills (e.g., standing, walking, and running) that can then be used to generate high-level behaviors. However, even with MoCap data, controlling simulated humanoids remains very hard, because this data offers only kinematic information. Finding physical control inputs to realize the MoCap-demonstrated motions has required methods like reinforcement learning that need large amounts of compute, which has effectively served as a barrier to entry for this exciting research direction.
In an effort to broaden participation and facilitate evaluation of ideas in humanoid locomotion research, we are releasing MoCapAct (Motion Capture with Actions), a library of high-quality pre-trained agents that can track over three hours of MoCap data for a simulated humanoid in the dm_control
physics-based environment and rollouts from these experts containing proprioceptive observations and actions. MoCapAct allows researchers to sidestep the computationally intensive task of training low-level control policies from MoCap data and instead use MoCapAct's expert agents and demonstrations for learning advanced locomotion behaviors. It also allows improving on our low-level policies by using them and their demonstration data as a starting point.
In our work, we use MoCapAct to train a single hierarchical policy capable of tracking the entire MoCap dataset within dm_control
.
We then re-use the learned low-level component to efficiently learn other high-level tasks.
Finally, we use MoCapAct to train an autoregressive GPT model and show that it can perform natural motion completion given a motion prompt.
We encourage the reader to visit our project website to see videos of our results as well as get links to our paper and code.
File Structure
The file structure of the dataset is:
βββ all
β βββ large
β β βββ large_1.tar.gz
β β βββ large_2.tar.gz
| β ...
β β βββ large_43.tar.gz
β βββ small
β βββ small_1.tar.gz
β βββ small_2.tar.gz
β βββ small_3.tar.gz
β
βββ sample
β βββ large.tar.gz
β βββ small.tar.gz
β
βββ videos
βββ full_clip_videos.tar.gz
βββ snippet_videos.tar.gz
MoCapAct Dataset Tarball Files
The dataset tarball files have the following structure:
all/small/small_*.tar.gz
: Contains HDF5 files with 20 rollouts per snippet. Due to file size limitations, we split the rollouts among multiple tarball files.all/large/large_*.tar.gz
: Contains HDF5 files with 200 rollouts per snippet. Due to file size limitations, we split the rollouts among multiple tarball files.sample/small.tar.gz
: Contains example HDF5 files with 20 rollouts per snippet.sample/large.tar.gz
: Contains example HDF5 files with 200 rollouts per snippet.
The HDF5 structure is detailed in Appendix A.2 of the paper as well as https://github.com/microsoft/MoCapAct#description.
An example for loading and inspecting an HDF5 file in Python is:
import h5py
dset = h5py.File("/path/to/small/CMU_083_33.hdf5", "r")
print("Expert actions from first rollout episode:")
print(dset["CMU_083_33-0-194/0/actions"][...])
MoCap Videos
There are two tarball files containing videos of the MoCap clips in the dataset:
full_clip_videos.tar.gz
contains videos of the full MoCap clips.snippet_videos.tar.gz
contains videos of the snippets that were used to train the experts. Note that they are playbacks of the clips themselves, not rollouts of the corresponding experts.
- Downloads last month
- 305