YanjieZe/3D-Diffusion-Policy
A universal visual imitation learning algorithm that uses 3D representations with diffusion models to train robotic manipulation policies.
Not currently ranked — collecting fresh signals.
star history
3D Diffusion Policy (DP3) combines 3D visual representations with diffusion policies to enable effective visuomotor control learning across diverse simulated and real-world robotics tasks. It handles both high-dimensional and low-dimensional control tasks with practical inference speed. The approach uses diffusion models to generate action sequences from visual observations, achieving generalizable robotic manipulation through learned policies.
Frequently asked
- What is YanjieZe/3D-Diffusion-Policy?
- A universal visual imitation learning algorithm that uses 3D representations with diffusion models to train robotic manipulation policies.
- Is 3D-Diffusion-Policy open source?
- Yes — YanjieZe/3D-Diffusion-Policy is open source, released under the MIT license.
- What language is 3D-Diffusion-Policy written in?
- YanjieZe/3D-Diffusion-Policy is primarily written in Python.
- How popular is 3D-Diffusion-Policy?
- YanjieZe/3D-Diffusion-Policy has 1.4k stars on GitHub.
- Where can I find 3D-Diffusion-Policy?
- YanjieZe/3D-Diffusion-Policy is on GitHub at https://github.com/YanjieZe/3D-Diffusion-Policy.