VoxWatch: An open-set speaker recognition benchmark on VoxCeleb

TL;DR

This paper introduces VoxWatch, the first public benchmark for open-set speaker recognition using VoxCeleb, analyzing how watchlist size and speech duration affect detection, and evaluating techniques like score calibration and fusion.

Contribution

It provides a systematic benchmark for OSI, quantifies the impact of watchlist size and speech duration, and evaluates the effectiveness of common techniques on this task.

Findings

Adaptive score normalization does not always improve OSI performance.

Score calibration significantly enhances detection accuracy.

Score fusion leads to notable performance improvements.

Abstract

Despite its broad practical applications such as in fraud prevention, open-set speaker identification (OSI) has received less attention in the speaker recognition community compared to speaker verification (SV). OSI deals with determining if a test speech sample belongs to a speaker from a set of pre-enrolled individuals (in-set) or if it is from an out-of-set speaker. In addition to the typical challenges associated with speech variability, OSI is prone to the "false-alarm problem"; as the size of the in-set speaker population (a.k.a watchlist) grows, the out-of-set scores become larger, leading to increased false alarm rates. This is in particular challenging for applications in financial institutions and border security where the watchlist size is typically of the order of several thousand speakers. Therefore, it is important to systematically quantify the false-alarm problem, and develop techniques that alleviate the impact of watchlist size on detection performance. Prior studies on this problem are sparse, and lack a common benchmark for systematic evaluations. In this paper, we present the first public benchmark for OSI, developed using the VoxCeleb dataset. We quantify the effect of the watchlist size and speech duration on the watchlist-based speaker detection task using three strong neural network based systems. In contrast to the findings from prior research, we show that the commonly adopted adaptive score normalization is not guaranteed to improve the performance for this task. On the other hand, we show that score calibration and score fusion, two other commonly used techniques in SV, result in significant improvements in OSI performance.