This is the official implementation of COMICS: End-to-end Bi-grained Contrastive Learning for Multi-face Forgery Detection
First install Detectron2 following the official guide: INSTALL.md.
Then build AdelaiDet with:
git clone https://github.com/zhangconghhh/COMICS.git
cd COMICS
python setup.py build develop
To train a model with COMICS, first setup the corresponding datasets following datasets/README.md, then run:
sh train_comics.sh
To evaluate the model after training, run:
sh eval_comics.sh
If this work is helpful to you, please cite it as:
@ARTICLE{comics24csvt,
title={COMICS: End-to-end Bi-grained Contrastive Learning for Multi-face Forgery Detection},
author={Cong Zhang, Honggang Qi, Shuhui Wang, Yuezun Li, Siwei Lyu},
journal={IEEE Transactions on Circuits and Systems for Video Technology},
year={2024},
volume={},
number={},
pages={},
doi={}
}
The codes are modified from AdelaiDet. Thanks for their open source.