New study examines spoofing attacks on voice biometrics across databases
16 May 2017 17:08 GMT

Researchers have completed a study that evaluated audio-based presentation attack detection across databases.

Authored by Pavel Korshunov and Sebastien Marcel from the Biometrics Group at Switzerland’s Idiap Research Institute, the paper states that the vulnerability of automatic speaker verification systems to spoofing or presentation attacks has limited their wide deployment.

Stating that it is important to develop mechanisms that can detect such attacks, the researchers added that it is equally important for these mechanisms to be seamlessly integrated into existing systems for practical and attack-resistant solutions.

Pointing to flaws in previous efforts, the team noted that while several audio-based presentation attack detection (PAD) methods have been proposed recently, that their evaluation was usually done on a single, often obscure, database with limited number of attacks.

“Therefore, in this paper, we conduct an extensive study of eight state-of-the-art PAD methods and evaluate their ability to detect known and unknown attacks (e.g., in a cross-database scenario) using two major publicly available speaker databases with spoofing attacks: AVspoof and ASVspoof“.

“We investigate whether combining several PAD systems via score fusion can improve attack detection accuracy. We also study the impact of fusing PAD systems (via parallel and cascading schemes) with two i-vector and inter-session variability based ASV systems on the overall performance in both bona fide (no attacks) and spoof scenarios”.

The authors state that the evaluation results question the efficiency and practicality of the existing PAD systems, especially when comparing results for individual databases and cross-database data.