English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

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

VoxWatch: An open-set speaker recognition benchmark on VoxCeleb