Anti-spooofing: Speaker verification vs. voice conversion

Voice conversion (VC) techniques, which modify a speaker’s voice to sound like another’s, present a threat to automatic speaker verification (SV) systems. In this paper, we evaluate the vulnerability of a state-of-the-art SV system against a converted speech spoofing attack. To overcome the spoofing attack, we implement state-of-the-art converted speech detectors based on shortand long-term features. We propose a new converted speech detector using a compact feature representation and a discriminative modeling approach. We experiment pairing converted speech detectors based on shortand long-term features to improve converted speech detection. The results indicate that the proposed fused converted speech detector outperforms state-of-the-art ones, achieving a detection accuracy of 98.1% for natural utterances and 97.6% for converted utterances. We include the anti-spoofing mechanism in our SV system as a post-processing module for accepted trials and reevaluate its performance, comparing it with the performance of an ideal system. Our results show that the SV system’s performance returns to acceptable values, with an average of 1.9% of the rejected trails remaining misclassified.