An Empirical Study on Channel Effects for Synthetic Voice Spoofing Countermeasure Systems
This addresses a practical issue for speaker verification systems by mitigating channel effects, though it is incremental as it builds on existing methods.
The paper tackled the problem of channel mismatch degrading spoofing countermeasure systems in speaker verification, showing significant performance drops in cross-dataset tests and proposing channel robust strategies that led to notable improvements.
Spoofing countermeasure (CM) systems are critical in speaker verification; they aim to discern spoofing attacks from bona fide speech trials. In practice, however, acoustic condition variability in speech utterances may significantly degrade the performance of CM systems. In this paper, we conduct a cross-dataset study on several state-of-the-art CM systems and observe significant performance degradation compared with their single-dataset performance. Observing differences of average magnitude spectra of bona fide utterances across the datasets, we hypothesize that channel mismatch among these datasets is one important reason. We then verify it by demonstrating a similar degradation of CM systems trained on original but evaluated on channel-shifted data. Finally, we propose several channel robust strategies (data augmentation, multi-task learning, adversarial learning) for CM systems, and observe a significant performance improvement on cross-dataset experiments.