PVRNN for human motion prediction for this paper: https://arxiv.org/abs/1906.06514

download data from https://github.com/una-dinosauria/human-motion-prediction

The two open source https://github.com/facebookresearch/QuaterNet

https://github.com/chaneyddtt/Convolutional-Sequence-to-Sequence-Model-for-Human-Dynamics are also very helpful

Run main.py with the following setting: --residual: True --veloc: True --loss_type: 1 --pos_embed: True

Test well with python3.6, Pytorch. Run with one GPU.

The visualization of results are shown in https://www.youtube.com/watch?v=Gmvk8HsyzOc&t=5s and https://www.youtube.com/watch?v=mECybzziYHM

If you are using this code, please cite this paper:

@article{wang2021pvred, title={PVRED: A Position-Velocity Recurrent Encoder-Decoder for Human Motion Prediction}, author={Wang, Hongsong and Dong, Jian and Cheng, Bin and Feng, Jiashi}, journal={IEEE Transactions on Image Processing}, year={2021}, publisher={IEEE} }