Views
No views yet
| Method | AP@0.5 | AP@0.75 | AP@0.95 | AR@100 | AR@50 | AR@20 | AR@10 |
|---|---|---|---|---|---|---|---|
| BA-TFD | 79.15 | 38.57 | 00.24 | 67.03 | 64.18 | 60.89 | 58.51 |
| BA-TFD+ | 96.30 | 84.96 | 04.44 | 81.62 | 80.48 | 79.40 | 78.75 |
pip install -r requirements.txt1python train.py \
2 --config ./config/batfd_default.toml \
3 --data_root <DATASET_PATH> \
4 --batch_size 4 --num_workers 8 --gpus 1 --precision 16ckpt directory, and the tensorboard log will be saved in lighntning_logs directory.1python train.py \
2 --config ./config/batfd_plus_default.toml \
3 --data_root <DATASET_PATH> \
4 --batch_size 4 --num_workers 8 --gpus 2 --precision 32FP32 for training BA-TFD+ as FP16 will cause inf and nan.ckpt directory, and the tensorboard log will be saved in lighntning_logs directory.ckpt directory.1python evaluate.py \
2 --config <CONFIG_PATH> \
3 --data_root <DATASET_PATH> \
4 --checkpoint <CHECKPOINT_PATH> \
5 --batch_size 1 --num_workers 4output directory, and the AP and AR scores will
be printed in the console.1@inproceedings{cai2022you,
2 title = {Do You Really Mean That? Content Driven Audio-Visual Deepfake Dataset and Multimodal Method for Temporal Forgery Localization},
3 author = {Cai, Zhixi and Stefanov, Kalin and Dhall, Abhinav and Hayat, Munawar},
4 booktitle = {2022 International Conference on Digital Image Computing: Techniques and Applications (DICTA)},
5 year = {2022},
6 doi = {10.1109/DICTA56598.2022.10034605},
7 pages = {1--10},
8 address = {Sydney, Australia},
9}1@article{cai2023glitch,
2 title = {Glitch in the Matrix: A Large Scale Benchmark for Content Driven Audio-Visual Forgery Detection and Localization},
3 author = {Cai, Zhixi and Ghosh, Shreya and Dhall, Abhinav and Gedeon, Tom and Stefanov, Kalin and Hayat, Munawar},
4 journal = {Computer Vision and Image Understanding},
5 year = {2023},
6 volume = {236},
7 pages = {103818},
8 issn = {1077-3142},
9 doi = {10.1016/j.cviu.2023.103818},
10}