Skip to content

Latest commit

 

History

History
11 lines (10 loc) · 507 Bytes

README.md

File metadata and controls

11 lines (10 loc) · 507 Bytes

dmp_primitives_pytorch

Tools for using motion primitives like Dynamic Motion Primitives or Differentiable Linear Dynamic Systems in PyTorch.

TODOs

  • Code the DMP part in a interpretable way.
  • Batched CEM optimization for DMPs
  • Code differentiable Linear Dynamic Systems (dLDS).
  • Code Cubic Splines
  • CEM optimization for dLDS and cubic splines.
  • Neural output code for training dLDS, splines and DMPs.
  • A very simple colab tutorial for introducing these primitives.