Attention-based latent features for jointly trained end-to-end automatic speech recognition with modified speech enhancement

In this paper, we propose a joint training framework that efficiently combines time-domain speech enhancement (SE) with an end-to-end (E2E) automatic speech recognition (ASR) system utilizing attention-based latent features. Using the latent feature to train E2E ASR implies that various time-domain...

ver descrição completa

Detalhes bibliográficos

Principais autores:	Da-Hee Yang, Joon-Hyuk Chang
Formato:	Artigo
Idioma:	Inglês
Publicado em:	Springer 2023-03-01
coleção:	Journal of King Saud University: Computer and Information Sciences
Assuntos:	Time-domain speech enhancement End-to-end automatic speech recognition Attention-based latent feature Joint training framework
Acesso em linha:	http://www.sciencedirect.com/science/article/pii/S1319157823000368
Tags:	Adicionar Tag Sem tags, seja o primeiro a adicionar uma tag!