Generalizable Audio Deepfake Detection via Latent Space Refinement and Augmentation
Authors:
Top Cited Papers
- 6 April 2025
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
Abstract
Advances in speech synthesis technologies, like text-to-speech (TTS) and voice conversion (VC), have made detecting deepfake speech increasingly challenging. Spoofing countermeasures often struggle to generalize effectively, particularly when faced with unseen attacks. To address this, we propose a novel strategy that integrates Latent Space Refinement (LSR) and Latent Space Augmentation (LSA) to improve the generalization of deepfake detection systems. LSR introduces multiple learnable prototypes for the spoof class, refining the latent space to better capture the intricate variations within spoofed data. LSA further diversifies spoofed data representations by applying augmentation techniques directly in the latent space, enabling the model to learn a broader range of spoofing patterns. We evaluated our approach on four representative datasets, i.e. ASVspoof 2019 LA, ASVspoof 2021 LA and DF, and In-The-Wild. The results show that LSR and LSA perform well individually, and their integration achieves competitive results, matching or surpassing current state-of-the-art methods.Keywords
Funding Information
- National Natural Science Foundation of China
This publication has 28 references indexed in Scilit:
- Transcending Forgery Specificity with Latent Space Augmentation for Generalizable Deepfake Detection2024
- CPAUG: Refining Copy-Paste Augmentation for Speech Anti-Spoofing2024
- One-Class Knowledge Distillation for Spoofing Speech Detection2024
- Can Large-Scale Vocoded Spoofed Data Improve Speech Spoofing Countermeasure with a Self-Supervised Front End?2024
- Audio Deepfake Detection With Self-Supervised Wavlm And Multi-Fusion Attentive Classifier2024
- AASIST: Audio Anti-Spoofing Using Integrated Spectro-Temporal Graph Attention Networks2022
- Rawboost: A Raw Data Boosting and Augmentation Method Applied to Automatic Speaker Verification Anti-Spoofing2022
- ASVspoof 2019: A large-scale public database of synthesized, converted and replayed speechComputer Speech & Language, 2020
- ArcFace: Additive Angular Margin Loss for Deep Face Recognition2019
- Empirical Risk Minimization2017