Inverse Filtering and Principal Component Analysis Techniques for Speech Dereverberation

Mohamed Anouar Ben Messaoud, Aicha Bouzid

Inverse Filtering and Principal Component Analysis Techniques for Speech Dereverberation

Číslo: 2/2016
Periodikum: Advances in Electrical and Electronic Engineering
DOI: 10.15598/aeee.v14i2.1533

Klíčová slova: Inverse filtering; kernel algorithm; principal components analysis; reverberant speech, Inverzní filtrování; Algoritmus jádra; Analýza hlavních komponent; Zřetelná řeč

Pro získání musíte mít účet v Citace PRO.

Přečíst po přihlášení

Anotace: In this work, we present a single channel approach for early and late reverberation suppression. This approach can be decomposed into two stages. The first stage employs the inverse filter to augment the signal-to-reverberant energy ratio. The second stage uses the kernel PCA algorithm to enhance the obtained dereverberant signal. It consists in extracting the main non-linear features from the speech signal after inverse filtering. Our approach appears to be efficient mainly in far field conditions and in highly reverberant environments.