Neural Motion Learner
This work is to extract skeletal structure from volumetric observations and to learn motion dynamics from the detected skeletal motions in a fully unsupervised manner.
Our model conducts motion generation/interpolation/retargeting based on the learned latent dynamics.
Note that it is an unofficial version of the work so that minimal amounts of codes are provided to demonstrate results.
Full descriptions including title, training codes and data pre-processing methods will be uploaded once the paper of this work is accepted to the conference.
We tested on Python 3.8 and Ubuntu 18.04 LTS.
The architecture is built from Pytorch 1.7.1 with Cuda 11.0.
Creating a conda environment is recommended.
## Download the repository git clone https://github.com/jinseokbae/neural_motion_learner.git cd neural_motion_learner ## Create conda env conda create --name nmotion python=3.8 conda activate nmotion ## modify setup.sh to match your cuda setting bash setup.sh
Using provided pretrained model, run demo codes to visualize followings:
## Motion generation python vis_generation.py ## Result will be stored in output/generation
## Motion interpolation python vis_interpolation.py ## Result will be stored in output/interpolation
## Motion retargeting python vis_retarget.py ## Result will be stored in output/retarget