Audio Deepfake Detection with Self-Supervised WavLM and Multi-Fusion Attentive Classifier

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Guo, Yinlin, Huang, Haofan, Chen, Xi, Zhao, He, Wang, Yuehai
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914636415107072
author Guo, Yinlin
Huang, Haofan
Chen, Xi
Zhao, He
Wang, Yuehai
author_facet Guo, Yinlin
Huang, Haofan
Chen, Xi
Zhao, He
Wang, Yuehai
contents With the rapid development of speech synthesis and voice conversion technologies, Audio Deepfake has become a serious threat to the Automatic Speaker Verification (ASV) system. Numerous countermeasures are proposed to detect this type of attack. In this paper, we report our efforts to combine the self-supervised WavLM model and Multi-Fusion Attentive classifier for audio deepfake detection. Our method exploits the WavLM model to extract features that are more conducive to spoofing detection for the first time. Then, we propose a novel Multi-Fusion Attentive (MFA) classifier based on the Attentive Statistics Pooling (ASP) layer. The MFA captures the complementary information of audio features at both time and layer levels. Experiments demonstrate that our methods achieve state-of-the-art results on the ASVspoof 2021 DF set and provide competitive results on the ASVspoof 2019 and 2021 LA set.
format Preprint
id arxiv_https___arxiv_org_abs_2312_08089
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Audio Deepfake Detection with Self-Supervised WavLM and Multi-Fusion Attentive Classifier
Guo, Yinlin
Huang, Haofan
Chen, Xi
Zhao, He
Wang, Yuehai
Audio and Speech Processing
With the rapid development of speech synthesis and voice conversion technologies, Audio Deepfake has become a serious threat to the Automatic Speaker Verification (ASV) system. Numerous countermeasures are proposed to detect this type of attack. In this paper, we report our efforts to combine the self-supervised WavLM model and Multi-Fusion Attentive classifier for audio deepfake detection. Our method exploits the WavLM model to extract features that are more conducive to spoofing detection for the first time. Then, we propose a novel Multi-Fusion Attentive (MFA) classifier based on the Attentive Statistics Pooling (ASP) layer. The MFA captures the complementary information of audio features at both time and layer levels. Experiments demonstrate that our methods achieve state-of-the-art results on the ASVspoof 2021 DF set and provide competitive results on the ASVspoof 2019 and 2021 LA set.
title Audio Deepfake Detection with Self-Supervised WavLM and Multi-Fusion Attentive Classifier
topic Audio and Speech Processing
url https://arxiv.org/abs/2312.08089