Accepted at ACM Multimedia 2025, Dublin, Ireland

SpecXNet: A Dual-Domain Convolutional Network for Robust Deepfake Detection [Paper]

Inzamamul Aalm, Md Tanvir Islam , Simon S. Woo^*,

| Sungkyunkwan University, Republic of Korea | *Corresponding Author |

---

Quick starts

Requirements

• pip install -r requirements.txt

Data preparation

You can follow the Pytorch implementation: https://github.com/pytorch/examples/tree/master/imagenet

Training

To train a model, run main.py with the desired model architecture and other super-paremeters:

 python3 main.py -a ffc_xception --lfu -data/csv [data/path] --use_se --gpu [gpu no]

Testing

Citation

If you find this work or code is helpful in your research, please cite:

@inproceedings{10.1145/3746027.3755707,
author = {Alam, Inzamamul and Islam, Md Tanvir and Woo, Simon S.},
title = {SpecXNet: A Dual-Domain Convolutional Network for Robust Deepfake Detection},
year = {2025},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
doi = {10.1145/3746027.3755707},
pages = {11667–11676},
numpages = {10},
location = {Dublin, Ireland},
series = {MM '25}
}

✅ To-Do (click to expand)

• Add FULL Train Code
• Add Test Code
• Add Model Weight