Learning Real Facial Concepts for Independent Deepfake Detection

2025-05-07

AI Summary

Deepfake detection models exhibit poor generalization under cross-domain settings, frequently misclassifying genuine faces as forged—primarily due to over-reliance on artifact-based cues and insufficient modeling of intrinsic characteristics of authentic human faces. To address this, we propose RealID, the first framework that explicitly decouples authentic face representation learning from artifact detection via a Real Concept Capture module (RealC2) and an Independent Dual Classifier (IDC). RealID incorporates a multi-prototype authentic memory bank, a dual-path decision mechanism, authenticity-guided contrastive learning, and probabilistic decoupled classification to effectively suppress non-artifact-related interference. Evaluated on five benchmark datasets, RealID achieves an average accuracy improvement of 1.74% over state-of-the-art methods and demonstrates significantly enhanced robustness against unseen deepfake generation techniques.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Deepfake detection models often struggle with generalization to unseen datasets, manifesting as misclassifying real instances as fake in target domains. This is primarily due to an overreliance on forgery artifacts and a limited understanding of real faces. To address this challenge, we propose a novel approach RealID to enhance generalization by learning a comprehensive concept of real faces while assessing the probabilities of belonging to the real and fake classes independently. RealID comprises two key modules: the Real Concept Capture Module (RealC2) and the Independent Dual-Decision Classifier (IDC). With the assistance of a MultiReal Memory, RealC2 maintains various prototypes for real faces, allowing the model to capture a comprehensive concept of real class. Meanwhile, IDC redefines the classification strategy by making independent decisions based on the concept of the real class and the presence of forgery artifacts. Through the combined effect of the above modules, the influence of forgery-irrelevant patterns is alleviated, and extensive experiments on five widely used datasets demonstrate that RealID significantly outperforms existing state-of-the-art methods, achieving a 1.74% improvement in average accuracy.

Problem

• Improving generalization in deepfake detection models
• Reducing misclassification of real instances as fake
• Learning comprehensive real face concepts independently

Innovation

• Learning comprehensive real face concepts
• Independent dual-decision classification strategy
• MultiReal Memory for diverse real prototypes