4 datasets found

The UASPEECH dataset is a dataset of speech recordings.

The TIMIT acoustic-phonetic continuous speech corpusCD-ROM contains a large collection of speech samples from 250 male and 250 female speakers.

The LJ-Speech dataset is a dataset of speech recordings of a female speaker.

Speaker verification systems experience significant performance degradation when tasked with short-duration trial recordings. To address this challenge, a multi-scale feature...