# GFADE: generalized feature adaptation and discrimination enhancement for deepfake detection

## TL;DR

This paper introduces GFADE, a new deepfake detection framework that improves performance across different datasets and manipulation types.

## Contribution

The novel framework combines multiple loss functions and the MixStyle technique to enhance generalization and robustness in deepfake detection.

## Key findings

- The model achieves superior detection accuracy in cross-dataset and cross-manipulation tests.
- Integration of multiple loss functions improves discriminative power and handles data imbalance effectively.
- MixStyle enhances generalization by introducing diverse visual styles during training.

## Abstract

With the rapid advancement of deep generative techniques, such as generative adversarial networks (GANs), the creation of realistic fake images and videos has become increasingly accessible, raising significant security and privacy concerns. Although existing deepfake detection methods perform well within a single dataset, they often experience substantial performance degradation when applied across datasets or manipulation types. To address this challenge, we propose a novel deepfake detection framework that combines multiple loss functions and the MixStyle technique. By integrating Cross-Entropy Loss, ArcFace loss, and Focal Loss, our model enhances its discriminative power to better handle complex forgery characteristics and effectively mitigate data imbalance. Additionally, the MixStyle technique introduces diverse visual styles during training, further improving the model’s generalization across different datasets and manipulation scenarios. Experimental results demonstrate that our method achieves superior detection accuracy across a range of cross-dataset and cross-manipulation tests, significantly improving model robustness and generalizability.