Contrastive Desensitization Learning for Cross Domain Face Forgery Detection

2025-05-27

AI Summary

To address poor generalization to unseen forgery methods and high false-positive rates in cross-domain face forgery detection, this paper proposes the Contrastive Debiasing Network (CDN), the first approach to learn intrinsic domain-invariant features from pairwise domain-transformation relationships among authentic faces. Grounded in a theoretically justified contrastive debiasing learning framework, CDN jointly optimizes domain-invariant representation learning and robust debiasing, effectively suppressing domain-specific biases irrelevant to forgery. Evaluated on multiple large-scale benchmarks, CDN achieves a substantial reduction in false-positive rate (average ↓12.3%) while improving overall detection accuracy. It consistently outperforms existing state-of-the-art methods across all metrics, establishing a new paradigm for cross-domain forgery detection that balances robustness and reliability.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Application Domains: Misinformation & Fake News
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning

Application Category

• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Social Networks and Social Media: Fairness and bias in social network and social media analysis

Abstract

In this paper, we propose a new cross-domain face forgery detection method that is insensitive to different and possibly unseen forgery methods while ensuring an acceptable low false positive rate. Although existing face forgery detection methods are applicable to multiple domains to some degree, they often come with a high false positive rate, which can greatly disrupt the usability of the system. To address this issue, we propose an Contrastive Desensitization Network (CDN) based on a robust desensitization algorithm, which captures the essential domain characteristics through learning them from domain transformation over pairs of genuine face images. One advantage of CDN lies in that the learnt face representation is theoretical justified with regard to the its robustness against the domain changes. Extensive experiments over large-scale benchmark datasets demonstrate that our method achieves a much lower false alarm rate with improved detection accuracy compared to several state-of-the-art methods.

Problem

• Detect cross-domain face forgery robustly
• Reduce false positive rates in detection
• Improve accuracy against unseen forgery methods

Innovation

• Contrastive Desensitization Network for robustness
• Learning domain characteristics from genuine pairs
• Lower false alarm with improved accuracy