Title: De-noising by Gammachirp and Wiener filter-based methods for speech enhancement

Authors: Hajer Rahali; Zied Hajaiej; Noureddine Ellouze

Addresses: Laboratory of Signal, Image and Information Technologies, BP 37, Le Belvédère, 1002 Tunis, Tunisie ' Laboratory of Signal, Image and Information Technologies, BP 37, Le Belvédère, 1002 Tunis, Tunisie ' Laboratory of Signal, Image and Information Technologies, BP 37, Le Belvédère, 1002 Tunis, Tunisie

Abstract: In this paper, we propose a method for enhancing of speech corrupted by noise. The new speech enhancement approach combines RASTA, Wiener filter (WF) and the Gammachirp filter (GF) in series connection to construct a two-stage hybrid system (named RASTA-WF-GF) in frequency domain to enhance the speech with additive noise. It is shown that the proposed method significantly outperforms, spectral subtraction (SS), Wiener filter (WF), Kalman filter (KF) and RASTA speech enhancement methods, in the presence of noise.

Keywords: Gammachirp filter; Wiener filter; robust speech recognition; noise reduction.

DOI: 10.1504/IJICT.2018.090416

International Journal of Information and Communication Technology, 2018 Vol.13 No.1, pp.55 - 67

Received: 14 Nov 2014
Accepted: 06 May 2015
Published online: 19 Mar 2018 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article