Data Quality as Predictor of Voice Anti-Spoofing Generalization - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2021

Data Quality as Predictor of Voice Anti-Spoofing Generalization

Abstract

Voice anti-spoofing aims at classifying a given utterance either as a bonafide human sample, or a spoofing attack (e.g. synthetic or replayed sample). Many anti-spoofing methods have been proposed but most of them fail to generalize across domains (corpora)-and we do not know why. We outline a novel interpretative framework for gauging the impact of data quality upon anti-spoofing performance. Our within-and betweendomain experiments pool data from seven public corpora and three anti-spoofing methods based on Gaussian mixture and convolutive neural network models. We assess the impacts of long-term spectral information, speaker population (through xvector speaker embeddings), signal-to-noise ratio, and selected voice quality features.
Fichier principal
Vignette du fichier
IS21_SpoofingDataQuality.pdf (354.76 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-03261131 , version 1 (15-06-2021)

Identifiers

Cite

Bhusan Chettri, Rosa González Hautamäki, Md Sahidullah, Tomi Kinnunen. Data Quality as Predictor of Voice Anti-Spoofing Generalization. INTERSPEECH 2021, Aug 2021, Brno, Czech Republic. ⟨10.21437/Interspeech.2021-1180⟩. ⟨hal-03261131⟩
98 View
144 Download

Altmetric

Share

Gmail Mastodon Facebook X LinkedIn More