Forensics Forest: Multi-scale Hierarchical Cascade Forest for Detecting GAN-generated Faces
Lists
Only you see this
Abstract
We describe a simple and effective method called ForensicsForest to detect GAN-generate faces. Instead of using the commonly used CNN models, we describe a novel multi-scale hierarchical cascade forest, which takes semantic and frequency features as input, and hierarchically cascades different levels of features for authenticity prediction. We then propose a multi-scale ensemble, which comprehensively considers different levels of information to improve the performance further. Our method is validated on state-of-the-art GAN-generated face datasets in comparison with several CNN models, which demonstrates the surprising effectiveness of our method in detecting GAN-generated faces.
BibTeX
@inproceedings{lu2023forensics,
author = {Jiucui Lu and Yuezun Li and Jiaran Zhou and Bin Li and Siwei Lyu},
title = {{Forensics Forest: Multi-scale Hierarchical Cascade Forest for Detecting GAN-generated Faces}},
booktitle = {ICME},
year = {2023},
doi = {10.1109/icme55011.2023.00394},
}Loading related content...