Conditional Neural Movement Primitives

This branch includes PyTorch implementation. For Keras/Tensorflow implementation please refer to Keras branch

This repository includes the CNMP implementation and example material from the experiments of the paper:

M. Yunus Seker, Mert Imre, Justus Piater, Emre Ugur, Conditional Neural Movement Primitives, RSS 2019

What's New ?

• PyTorch implementation of the framework with:
  • Multi-step/whole-trajectory prediction support
  • Easier data/model organization with a more basic batch-based approach
  • Recreated Experiment 1 from the paper

Requirements:

• Python >=3.6

• Jupyter Notebook

• PyTorch >=1.6.0

Contact Information:

• M. Yunus Seker , m.yunusseker@hotmail.com, yunus.seker1@boun.edu.tr

• Mert Imre , mert.imre@boun.edu.tr

• Emre Ugur, emre.ugur@boun.edur.tr