Optimizing a-DCF for Spoofing-Robust Speaker Verification

TL;DR

This paper introduces an optimized spoofing-robust speaker verification system using a-DCF, achieving significant improvements in detection cost function over previous methods by combining a-DCF with BCE and novel thresholding.

Contribution

It presents a new method that directly optimizes a-DCF for spoofing-robust speaker verification, integrating threshold optimization and fusion techniques for enhanced performance.

Findings

13% relative improvement over BCE-only system

43% relative improvement with non-linear score fusion

Significant reduction in minimum a-DCF scores

Abstract

Automatic speaker verification (ASV) systems are vulnerable to spoofing attacks. We propose a spoofing-robust ASV system optimized directly for the recently introduced architecture-agnostic detection cost function (a-DCF), which allows targeting a desired trade-off between the contradicting aims of user convenience and robustness to spoofing. We combine a-DCF and binary cross-entropy (BCE) with a novel straightforward threshold optimization technique. Our results with an embedding fusion system on ASVspoof2019 data demonstrate relative improvement of $13\%$ over a system trained using BCE only (from minimum a-DCF of $0.1445$ to $0.1254$). Using an alternative non-linear score fusion approach provides relative improvement of $43\%$ (from minimum a-DCF of $0.0508$ to $0.0289$).