/Spatio-Temporal-Perturbations-for-Video-Attribution

The source code for the journal paper: Spatio-Temporal Perturbations for Video Attribution, TCSVT-2021

Primary LanguagePython

Spatio-Temporal Perturbations for Video Attribution

Code for paper on TCSVT 2021: Spatio-Temporal Perturbations for Video Attribution, which is an extension paper of our previous work on WACV 2021: Towards Visually Explaining Video Understanding Networks with Perturbation.

Overview

The project is build upon Github repo of Video-Visual-Explanations.

More details regarding the code will be elaborated soon.

Methodology

Reference

TCSVT 2021

@article{VideoAttribution,
  title={Spatio-Temporal Perturbations for Video Attribution}, 
  author={Li, Zhenqiang and Wang, Weimin and Li, Zuoyue and Huang, Yifei and Sato, Yoichi},
  journal={IEEE Transactions on Circuits and Systems for Video Technology}, 
  year={2021},
  doi={10.1109/TCSVT.2021.3081761}}

WACV 2021

@article{li2020comprehensive,
  title={A Comprehensive Study on Visual Explanations for Spatio-temporal Networks},
  author={Li, Zhenqiang and Wang, Weimin and Li, Zuoyue and Huang, Yifei and Sato, Yoichi},
  journal={arXiv preprint arXiv:2005.00375},
  year={2020}
}