End-to-End Speech Recognition With Joint Dereverberation Of Sub-Band Autoregressive Envelopes

The end-to-end (E2E) automatic speech recognition (ASR) systems are often required to operate in reverberant conditions, where the long-term sub-band envelopes of the speech are temporally smeared. In this paper, we develop a feature enhancement approach using a neural model operating on sub-band te...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	arXiv.org 2022-02
Hauptverfasser:	Kumar, Rohit, Purushothaman, Anurenjan, Sreeram, Anirudh, Ganapathy, Sriram
Format:	Artikel
Sprache:	eng
Schlagworte:	Automatic speech recognition Cost function Datasets Envelopes Feature extraction Linear prediction Modelling Neural networks Optimization Speech Voice recognition
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!