Anti-Spoofing Using Transfer Learning with Variational Information Bottleneck
This work addresses the threat of synthetic speech spoofing to speaker verification systems, offering an incremental improvement with enhanced generalization.
The authors tackled the problem of distinguishing synthetic spoofed speech from genuine speech in automatic speaker verification systems, proposing a transfer learning method with variational information bottleneck that outperforms state-of-the-art systems on the ASVspoof 2019 LA database and shows robustness in low-resource and cross-dataset settings.
Recent advances in sophisticated synthetic speech generated from text-to-speech (TTS) or voice conversion (VC) systems cause threats to the existing automatic speaker verification (ASV) systems. Since such synthetic speech is generated from diverse algorithms, generalization ability with using limited training data is indispensable for a robust anti-spoofing system. In this work, we propose a transfer learning scheme based on the wav2vec 2.0 pretrained model with variational information bottleneck (VIB) for speech anti-spoofing task. Evaluation on the ASVspoof 2019 logical access (LA) database shows that our method improves the performance of distinguishing unseen spoofed and genuine speech, outperforming current state-of-the-art anti-spoofing systems. Furthermore, we show that the proposed system improves performance in low-resource and cross-dataset settings of anti-spoofing task significantly, demonstrating that our system is also robust in terms of data size and data distribution.