Comparison of Modern Deep Learning Models for Speaker Verification

Comparison of Modern Deep Learning Models for Speaker Verification

This research presents an extensive comparative analysis of a selection of popular deep speaker embedding models, namely WavLM, TitaNet, ECAPA, and PyAnnote, applied in speaker verification tasks. The study employs a specially curated dataset, specifically designed to mirror the real-world operating...

Full description

Bibliographic Details
Main Authors: Vitalii Brydinskyi, Yuriy Khoma, Dmytro Sabodashko, Michal Podpora, Volodymyr Khoma, Alexander Konovalov, Maryna Kostiak
Format: Article
Language:English
Published: MDPI AG 2024-02-01
Series:Applied Sciences
Subjects:
speaker embedding models
speaker verification
non-English speech dataset evaluation
Online Access:https://www.mdpi.com/2076-3417/14/4/1329

Internet

https://www.mdpi.com/2076-3417/14/4/1329

Similar Items