Truncate-Split-Contrast: A Framework for Learning from Mislabeled Videos [arXiv]

@article{zixiao2022truncate,
  title={Truncate-Split-Contrast: A Framework for Learning from Mislabeled Videos},
  author={Zixiao, Wang and Junwu, Weng and Chun, Yuan and Jue, Wang},
  journal={arXiv preprint arXiv:2212.13495},
  year={2022}
}

Overview

We release the PyTorch code of the Truncate-Split-Contrast.

Content

TODO

Prerequisites

TODO

Data Preparation

TODO

Code

This code is based on the TSM codebase.

Training and Testing

You can train and test by one line bash start.sh

You may need to modify the file opts.py and train.sh to fit your environment.

Important

This is a very early version of our code. We will release the full version later.