SFE-Net: Harnessing Biological Principles of Differential Gene Expression for Improved Feature Selection in Deep Learning Networks

TL;DR

The paper tackles the challenge of DeepFake detection under varying synthesis methods by addressing feature representation drift across datasets. It introduces Selective Feature Expression Network (SFE-Net), a bio-inspired framework that dynamically modulates feature priorities in response to input characteristics, inspired by differential gene expression. Across FaceForensics++ and Celeb-DF datasets, SFE-Net achieves superior cross-dataset generalization, particularly in the frequency-domain detection setting, though performance still varies by dataset and comes with higher computational costs. The work demonstrates how adaptive feature selection can mitigate overfitting and improve robustness in digital forensics, paving the way for more responsive deep learning systems against evolving forgeries.

Abstract

In the realm of DeepFake detection, the challenge of adapting to various synthesis methodologies such as Faceswap, Deepfakes, Face2Face, and NeuralTextures significantly impacts the performance of traditional machine learning models. These models often suffer from static feature representation, which struggles to perform consistently across diversely generated deepfake datasets. Inspired by the biological concept of differential gene expression, where gene activation is dynamically regulated in response to environmental stimuli, we introduce the Selective Feature Expression Network (SFE-Net). This innovative framework integrates selective feature activation principles into deep learning architectures, allowing the model to dynamically adjust feature priorities in response to varying deepfake generation techniques. SFE-Net employs a novel mechanism that selectively enhances critical features essential for accurately detecting forgeries, while reducing the impact of irrelevant or misleading cues akin to adaptive evolutionary processes in nature. Through rigorous testing on a range of deepfake datasets, SFE-Net not only surpasses existing static models in detecting sophisticated forgeries but also shows enhanced generalization capabilities in cross-dataset scenarios. Our approach significantly mitigates overfitting by maintaining a dynamic balance between feature exploration and exploitation, thus producing more robust and effective deepfake detection models. This bio-inspired strategy paves the way for developing adaptive deep learning systems that are finely tuned to address the nuanced challenges posed by the varied nature of digital forgeries in modern digital forensics.

Figures (3)

• Figure 1: Introducing feature-selective expression based on gene-selective expression: Lico--light consistency feature, Hifr-- high frequency feature, Comr--compression reconstruction feature, Moop--morphological operation feature, Text--texture feature
• Figure 2: Visualizing the comparison of different feature extractions.
• Figure 3: SFE-Net architecture