Score: 0

UMCL: Unimodal-generated Multimodal Contrastive Learning for Cross-compression-rate Deepfake Detection

Published: November 24, 2025 | arXiv ID: 2511.18983v1

By: Ching-Yi Lai , Chih-Yu Jian , Pei-Cheng Chuang and more

Potential Business Impact:

Finds fake videos even when they are squeezed.

Business Areas:
Image Recognition Data and Analytics, Software

In deepfake detection, the varying degrees of compression employed by social media platforms pose significant challenges for model generalization and reliability. Although existing methods have progressed from single-modal to multimodal approaches, they face critical limitations: single-modal methods struggle with feature degradation under data compression in social media streaming, while multimodal approaches require expensive data collection and labeling and suffer from inconsistent modal quality or accessibility in real-world scenarios. To address these challenges, we propose a novel Unimodal-generated Multimodal Contrastive Learning (UMCL) framework for robust cross-compression-rate (CCR) deepfake detection. In the training stage, our approach transforms a single visual modality into three complementary features: compression-robust rPPG signals, temporal landmark dynamics, and semantic embeddings from pre-trained vision-language models. These features are explicitly aligned through an affinity-driven semantic alignment (ASA) strategy, which models inter-modal relationships through affinity matrices and optimizes their consistency through contrastive learning. Subsequently, our cross-quality similarity learning (CQSL) strategy enhances feature robustness across compression rates. Extensive experiments demonstrate that our method achieves superior performance across various compression rates and manipulation types, establishing a new benchmark for robust deepfake detection. Notably, our approach maintains high detection accuracy even when individual features degrade, while providing interpretable insights into feature relationships through explicit alignment.

Country of Origin
🇹🇼 Taiwan, Province of China

Page Count
24 pages

Category
Computer Science:
CV and Pattern Recognition