RPRA-ADD: Forgery Trace Enhancement-Driven Audio Deepfake Detection

2025-05-31

AI Summary

To address poor generalization and insufficient forgery-trace perception in deepfake audio detection, this paper proposes an end-to-end detection framework driven synergistically by reconstruction, perception, reinforcement, and attention. We innovatively design a global–local forgery perception module, a multi-stage discrete contrastive loss, and a dynamic attention mechanism that focuses on forgery traces via a reconstruction-difference matrix. The framework integrates a self-supervised reconstruction network with dual-scale feature modeling, significantly enhancing robustness against unseen spoofing techniques and cross-acoustic-domain scenarios. Evaluated on four major benchmarks—ASVspoof2019, ASVspoof2021, CodecFake, and FakeSound—the method achieves state-of-the-art performance, with average improvements exceeding 20% over prior works. In rigorous 3×3 cross-domain evaluations, it demonstrates superior robustness, outperforming all existing approaches.

Technology Category

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

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Security and Privacy: Cyber-crime defenses and forensics
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search

Abstract

Existing methods for deepfake audio detection have demonstrated some effectiveness. However, they still face challenges in generalizing to new forgery techniques and evolving attack patterns. This limitation mainly arises because the models rely heavily on the distribution of the training data and fail to learn a decision boundary that captures the essential characteristics of forgeries. Additionally, relying solely on a classification loss makes it difficult to capture the intrinsic differences between real and fake audio. In this paper, we propose the RPRA-ADD, an integrated Reconstruction-Perception-Reinforcement-Attention networks based forgery trace enhancement-driven robust audio deepfake detection framework. First, we propose a Global-Local Forgery Perception (GLFP) module for enhancing the acoustic perception capacity of forgery traces. To significantly reinforce the feature space distribution differences between real and fake audio, the Multi-stage Dispersed Enhancement Loss (MDEL) is designed, which implements a dispersal strategy in multi-stage feature spaces. Furthermore, in order to enhance feature awareness towards forgery traces, the Fake Trace Focused Attention (FTFA) mechanism is introduced to adjust attention weights dynamically according to the reconstruction discrepancy matrix. Visualization experiments not only demonstrate that FTFA improves attention to voice segments, but also enhance the generalization capability. Experimental results demonstrate that the proposed method achieves state-of-the-art performance on 4 benchmark datasets, including ASVspoof2019, ASVspoof2021, CodecFake, and FakeSound, achieving over 20% performance improvement. In addition, it outperforms existing methods in rigorous 3*3 cross-domain evaluations across Speech, Sound, and Singing, demonstrating strong generalization capability across diverse audio domains.

Problem

• Enhancing detection of evolving deepfake audio techniques
• Improving generalization across diverse forgery patterns
• Strengthening feature space differences for real vs fake audio

Innovation

• Global-Local Forgery Perception module enhances acoustic traces
• Multi-stage Dispersed Enhancement Loss reinforces feature differences
• Fake Trace Focused Attention adjusts weights dynamically