Divide and Conquer: Reliable Multi-View Evidential Learning for Deepfake Detection

2026-06-01

AI Summary

This work addresses the limitations of existing single-view deepfake detection methods, which often exhibit poor generalization and overconfident predictions due to semantic features obscuring subtle forgery traces. To mitigate this issue, the authors propose DiCoME, a novel framework that decouples semantic content from forgery-related artifacts through geometric projection, thereby alleviating semantic masking via geometry-aware view purification. Furthermore, DiCoME integrates multi-view information and leverages uncertainty-aware evidential learning combined with cognitive conflict modeling to produce well-calibrated uncertainty estimates. Extensive experiments demonstrate that the proposed method significantly outperforms state-of-the-art approaches across multiple benchmarks, achieving both enhanced generalization and reliable quantification of detection confidence.

Abstract

With the evolution of generative models, deepfakes have achieved near-perfect semantic realism, leaving forensic traces only in subtle structural anomalies. However, existing single-view paradigms often fail to generalize, as dominant semantic features overwhelm subtle artifact cues within entangled representations. This imbalance leads to overconfident yet brittle predictions -- a phenomenon we term the Semantic Masking Effect. To address this challenge, we propose a reliable framework called Divide-and-Conquer Multi-View Evidential Learning (DiCoME) for Deepfake Detection. In the "Divide" phase, we employ Geometric View Purification to decompose the entangled representation space through principled geometric projection. This process suppresses semantic interference within artifact-sensitive representations, forming the foundation for decorrelated yet complementary semantic and artifact views. In the "Conquer" phase, we leverage Uncertainty-Aware Evidential Learning to synthesize these distinct views. By explicitly modeling the "epistemic conflict" between semantic and artifact cues, this mechanism provides calibrated uncertainty estimates instead of forcing rigid deterministic decisions. Extensive experiments across multiple benchmarks demonstrate that our method consistently outperforms existing approaches in generalization performance, while providing reliable uncertainty estimation for trustworthy deepfake detection. Code is available at https://github.com/kxl0825/DiCoME.git.

Problem

• Deepfake Detection
• Semantic Masking Effect
• Multi-View Learning
• Generalization
• Uncertainty Estimation

Innovation

• Geometric View Purification
• Uncertainty-Aware Evidential Learning
• Semantic Masking Effect
• Multi-View Representation
• Epistemic Conflict