motion_imitationerwincoumans/motion_imitation
Enables quadruped robots to imitate reference motion trajectories via reinforcement learning,…
No ratings yet
Rate this skill
Stars1.5k
Forks315
Watchers1.5k
Open Issues15
InstallOne-line install
UnconfirmedWhether this repository follows the Agent Skills spec (that is, whether it contains a SKILL.md) has not been verified yet. Until that is settled, clone it and use what you need:
git clone https://github.com/erwincoumans/motion_imitation.git
OverviewOverview
Key Features
Enables quadruped robots to imitate reference motion trajectories via reinforcement learning, with a complete command-line workflow for training, testing, visualization and model saving. Key capabilities include: 1. Mode switching: the --mode parameter toggles between train and test, continuously optimizing the policy during training and loading an existing model for direct evaluation during testing; 2. Reference motion loading: the --motion_file parameter specifies a reference motion data file (by default read from motion_imitation/data/motions/) so the robot learns to imitate that trajectory; 3. Model persistence: the --model_file parameter sets the model file path, and --int_save_freq automatically saves checkpoints at fixed iteration intervals to avoid losing progress on interruption; 4. Training visualization: the --visualize flag enables graphical rendering to observe robot posture and imitation results in real time; 5. Parallel environment configuration: the -n parameter sets the number of parallel environments to increase sampling throughput and training efficiency.
Use Cases
Reinforcement learning training for quadruped gait and motion imitation, loading and playback validation of reference motion data, visual debugging of the training process, checkpoint management during long training runs, and policy testing/evaluation after training.
Target Roles
AI / Algorithm Engineer, Hardware / Embedded Engineer, Test / QA Engineer.
RatingUser ratings
Nobody has rated this skill yet. Yours would be the first on the site — your vote becomes the initial average.
RepositoryRepository
| Skill ID | 1642 |
|---|---|
| Data source | GitHub |
| Repository | erwincoumans/motion_imitation |
| Category | AI & Robotics / Robot Motion Control & Reinforcement Learning |
| Primary language | Python |
| License | Apache License 2.0 |
| Default branch | master |
| GitHub repository ID | 259765011 |
| Node ID | MDEwOlJlcG9zaXRvcnkyNTk3NjUwMTE= |
| Rank | #547 (position within the source ranking) |
| Site-wide popularity rank | #634 (computed live from Stars, across all 5833 entries) |
GrowthStar / Fork growth
One snapshot per day, tracking how Stars and Forks changed
- Now1,469
- Previous point · 09-23 1,469
- vs previous point ±0
- 2 data points 18 days
- Now315
- Previous point · 09-23 315
- vs previous point ±0
- 2 data points 18 days
TimelineTimeline
Repository created2020-04-28 22:23:002020-04-28
Data last updated2026-10-26 11:59:5523 days ago
Last commit2023-03-24 23:54:562023-03-24
CommitLatest commit
Committed by 匿名贡献者. Update Category: Documentation Updates. Commit details: Update README.md. This update refines the documentation for easier onboarding.