Skip to content

The Official PyTorch implementation of "Part Aware Contrastive Learning for Self-Supervised Action Recognition" in IJCAI 2023

Notifications You must be signed in to change notification settings

GitHubOfHyl97/SkeAttnCLR

Repository files navigation

SkeAttnCLR

The Official PyTorch implementation of "Part Aware Contrastive Learning for Self-Supervised Action Recognition" in IJCAI 2023. The arXiv version of our paper is in https://arxiv.org/abs/2305.00666.

Requirements

Python >= 3.6, Pytorch >= 1.4

Data Preparation

  • Download the raw data of NTU_RGB+D and PKU-MMD.
  • For NTU RGB+D dataset, preprocess data with tools/ntu_gendata.py. For PKU-MMD dataset, preprocess data with tools/pku_part1_gendata.py.
  • Then downsample the data to 64 frames with feeder/preprocess_ntu.py and feeder/preprocess_pku.py.

Pre-Training & Linear Evaluation

Example for pre-training and linear evaluation of SkeAttnCLR. You can change some settings of .yaml files in config/SkeAttnCLR/NTU60 folder.

sh run_NTU60.sh

Citation

Please cite our paper if you find this repository useful in your resesarch:

@inproceedings{Hua2023SkeAttnCLR,
  Title= {Part Aware Contrastive Learning for Self-Supervised Action Recognition},
  Author= {Hua, Yilei and Wu, Wenhan and Zheng, Ce and Lu, Aidong and Liu, Mengyuan and Chen, Chen and Wu, Shiqian},
  Booktitle= {International Joint Conference on Artificial Intelligence},
  Year= {2023}
}

About

The Official PyTorch implementation of "Part Aware Contrastive Learning for Self-Supervised Action Recognition" in IJCAI 2023

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published