Skip to content

About

Generate and learn collision free movement using a Riemannian approach

Resources

Stars

0 stars

Watchers

0 watching

Forks

 
 

Latest commit

 

History

456 Commits

Folders and files

Repository files navigation

(pyrieef) Python Riemannian Electrical Fields

Pure python project to generate and learn collision free movement in real environments using a Riemannian geometric approach.

Install

pip install -r requirements.txt

Run all tests

pytest tests

Result of

python examples/plot_barrier.py

EF

For a technical presentation of the algorithm developed in this work refer to the following publication

@inproceedings{mainprice2016warping,
  title={Warping the workspace geometry with electric potentials for motion optimization of manipulation tasks},
  author={Mainprice, Jim and Ratliff, Nathan and Schaal, Stefan},
  booktitle={Intelligent Robots and Systems (IROS), IEEE/RSJ International Conference on},
  pages={3156--3163},
  year={2016}
}

About

Generate and learn collision free movement using a Riemannian approach

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages