A Study On Data Augmentation In Voice Anti-Spoofing
This work addresses the challenge of enhancing anti-spoofing systems for voice authentication, which is crucial for securing speech-based applications, though it is incremental as it builds on existing data-centric approaches.
The paper tackled the problem of improving synthetic or spoofed audio detection by studying data augmentation techniques to handle channel variability, audio compressions, bandwidths, and unseen attacks, achieving state-of-the-art performance with an EER of 14.46% in the Deep Fake category and reducing the best baseline EER by 50% in the Logical Access task.
In this paper, we perform an in-depth study of how data augmentation techniques improve synthetic or spoofed audio detection. Specifically, we propose methods to deal with channel variability, different audio compressions, different band-widths, and unseen spoofing attacks, which have all been shown to significantly degrade the performance of audio-based systems and Anti-Spoofing systems. Our results are based on the ASVspoof 2021 challenge, in the Logical Access (LA) and Deep Fake (DF) categories. Our study is Data-Centric, meaning that the models are fixed and we significantly improve the results by making changes in the data. We introduce two forms of data augmentation - compression augmentation for the DF part, compression & channel augmentation for the LA part. In addition, a new type of online data augmentation, SpecAverage, is introduced in which the audio features are masked with their average value in order to improve generalization. Furthermore, we introduce a Log spectrogram feature design that improved the results. Our best single system and fusion scheme both achieve state-of-the-art performance in the DF category, with an EER of 15.46% and 14.46% respectively. Our best system for the LA task reduced the best baseline EER by 50% and the min t-DCF by 16%. Our techniques to deal with spoofed data from a wide variety of distributions can be replicated and can help anti-spoofing and speech-based systems enhance their results.