Can We Use Speaker Recognition Technology to Attack Itself? Enhancing Mimicry Attacks Using Automatic Target Speaker Selection
Abstract
We consider technology-assisted mimicry attacks in the context of automatic speaker verification (ASV). We use ASV itself to select targeted speakers to be attacked by human-based mimicry. We recorded 6 naive mimics for whom we select target celebrities from VoxCeleb1 and VoxCeleb2 corpora (7,365 potential targets) using an i-vector system. The attacker attempts to mimic the selected target, with the utterances subjected to ASV tests using an independently developed x-vector system. Our main finding is negative: even if some of the attacker scores against the target speakers were slightly increased, our mimics did not succeed in spoofing the x-vector system. Interestingly, however, the relative ordering of the selected targets (closest, furthest, median) are consistent between the systems, which suggests some level of transferability between the systems.
Domains
Computer Science [cs] Computer Vision and Pattern Recognition [cs.CV] Computer Science [cs] Artificial Intelligence [cs.AI] Computer Science [cs] Human-Computer Interaction [cs.HC] Computer Science [cs] Machine Learning [cs.LG] Computer Science [cs] Multimedia [cs.MM] Engineering Sciences [physics] Acoustics [physics.class-ph] Engineering Sciences [physics] Signal and Image processing Cognitive science Linguistics
Origin : Files produced by the author(s)
Loading...