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

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...

Full description

Bibliographic Details
Main Authors: Da-Hee Yang, Joon-Hyuk Chang
Format: Article
Language:English
Published: Elsevier 2023-03-01
Series:Journal of King Saud University: Computer and Information Sciences
Subjects:
Time-domain speech enhancement
End-to-end automatic speech recognition
Attention-based latent feature
Joint training framework
Online Access:http://www.sciencedirect.com/science/article/pii/S1319157823000368

Internet

http://www.sciencedirect.com/science/article/pii/S1319157823000368

Similar Items