Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Motion imitation with deep reinforcement learning.
| Date | Stars |
|---|---|
| 2026-07-31 | 3068 |
| 2026-08-01 | 3068 |
| 2026-08-02 | 3068 |
| 2026-08-03 | 3071 |
| 2026-08-04 | 3072 |
| 2026-08-05 | 3074 |
| 2026-08-06 | 3074 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# Intro **This codebase is now deprecated.** Checkout [MimicKit](https://github.com/xbpeng/MimicKit/) for much faster and easy-to-use implementations of our methods. Code accompanying the following papers: "DeepMimic: Example-Guided Deep Reinforcement Learning of Physics-Based Character Skills" \ (https://xbpeng.github.io/projects/DeepMimic/index.html) \  "AMP: Adversarial Motion Priors for Stylized Physics-Based Character Control" \ (https://xbpeng.github.io/projects/AMP/index.html) \  The framework uses reinforcement learning to train a simulated humanoid to imitate a variety of motion skills from mocap data. ## Dependencies ``sudo apt install libgl1-mesa-dev libx11-dev libxrandr-dev libxi-dev`` ``sudo apt install mesa-utils`` ``sudo apt install clang`` ``sudo apt install cmake`` C++: - Bullet 2.88 (https://github.com/bulletphysics/bullet3/releases) Download Bullet 2.88 from the above link and install using the following commands. ``./build_cmake_pybullet_double.sh`` ``cd build_cmake`` ``sudo make install`` - Eigen (http://www.eigen.tuxfamily.org/index.php?title=Main_Page) (Version : 3.3.7) ``mkdir build && cd build`` ``cmake ..`` ``sudo make install`` - OpenGL >= 3.2 - freeglut (http://freeglut.sourceforge.net/) ( Version : 3.0.0 ) ``cmake .`` ``make`` ``sudo make install`` - glew (http://glew.sourceforge.net/) ( Version : 2.1.0 ) ``make`` ``sudo make install`` ``make clean`` Misc: - SWIG (http://www.swig.org/) ( Version : 4.0.0 ) ``./configure --without-pcre`` ``make`` ``sudo make install`` - MPI - Windows: https://docs.microsoft.com/en-us/message-passing-interface/microsoft-mpi - Linux: `sudo apt install libopenmpi-dev` Python: - Python 3 - PyOpenGL (http://pyopengl.sourceforge.net/) ``pip install PyOpenGL PyOpenGL_accelerate`` - Tensorflow (https://www.tensorflow.org/) ( Vesrion : 1.13.1 ) ``pip install tensorflow`` - MPI4Py (https://mpi4py.readthedocs.io/en/stable/install.html) ``pip install mpi4py`` ## Build The simulated environments are written in C++, and the python wrapper is built using SWIG. Note that MPI must be installed before MPI4Py. When building Bullet, be sure to disable double precision with the build flag `USE_DOUBLE_PRECISION=OFF`. ### Windows The wrapper is built using `DeepMimicCore.sln`. 1. Select the `x64` configuration from the configuration manager. 2. Under the project properties for `DeepMimicCore` modify `Additional Include Directories` to specify - Bullet source directory - Eigen include directory - python include directory 3. Modify `Additional Library Directories` to specify - Bullet lib directory - python lib directory 4. Build `DeepMimicCore` project with the `Release_Swig` configuration and this should generate `DeepMimicCore.py` in `DeepMimicCore/`. ### Linux 1. Modify the `Makefile` in `DeepMimicCore/` by specifying the following, - `EIGEN_DIR`: Eigen include directory - `BULLET_INC_DIR`: Bullet source directory - `PYTHON_INC`: python include directory - `PYTHON_LIB`: python lib directory 2. Build wrapper, ``` make python ``` This should generate `DeepMimicCore.py` in `DeepMimicCore/` ## How to Use Once the python wrapper has been built, training is done entirely in python using Tensorflow. `DeepMimic.py` runs the visualizer used to view the simulation. Training is done with `mpi_run.py`, which uses MPI to parallelize training across multiple processes. `DeepMimic.py` is run by specifying an argument file that provides the configurations for a scene. For example, ``` python DeepMimic.py --arg_file args/run_humanoid3d_spinkick_args.txt ``` will run a pre-trained policy for a spinkick. Similarly, ``` python DeepMimic.py --arg_file args/play_motion_humanoid3d_args.txt ``` will load and play a mocap clip. To run a pre-trained policy for a simulated dog, use this command ``` python DeepMimic.py --arg_file args/run_dog3d
Excerpt of 8,318 characters
Read on GitHub
Simon Fraser University, NVIDIA
51
3
2
1
1
1
1
1
Logan King
1
William Tambellini · RWS · United States
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:1c7c347a4a9dfdfc, desc:reinforcement learning