Beyond Surface Artifacts: Capturing Shared Latent Forgery Knowledge Across Modalities

TL;DR

This paper introduces a modality-agnostic forgery detection framework that captures shared latent forgery knowledge across modalities, significantly improving generalization to unseen and dark modalities in multimodal deepfake detection.

Contribution

It proposes the first modality-agnostic forgery detection framework and a benchmark to evaluate cross-modal generalization, advancing universal multimodal forgery defense.

Findings

Empirically demonstrates the existence of universal forgery traces.

Achieves significant performance improvements on unknown modalities.

Introduces the DeepModal-Bench for rigorous generalization assessment.

Abstract

As generative artificial intelligence evolves, deepfake attacks have escalated from single-modality manipulations to complex, multimodal threats. Existing forensic techniques face a severe generalization bottleneck: by relying excessively on superficial, modality-specific artifacts, they neglect the shared latent forgery knowledge hidden beneath variable physical appearances. Consequently, these models suffer catastrophic performance degradation when confronted with unseen "dark modalities." To break this limitation, this paper introduces a paradigm shift that redefines multimodal forensics from conventional "feature fusion" to "modality generalization." We propose the first modality-agnostic forgery (MAF) detection framework. By explicitly decoupling modality-specific styles, MAF precisely extracts the essential, cross-modal latent forgery knowledge. Furthermore, we define two progressive dimensions to quantify model generalization: transferability toward semantically correlated modalities (Weak MAF), and robustness against completely isolated signals of "dark modality" (Strong MAF). To rigorously assess these generalization limits, we introduce the DeepModal-Bench benchmark, which integrates diverse multimodal forgery detection algorithms and adapts state-of-the-art generalized learning methods. This study not only empirically proves the existence of universal forgery traces but also achieves significant performance breakthroughs on unknown modalities via the MAF framework, offering a pioneering technical pathway for universal multimodal defense.