Role of Noise Estimation in Enhancement of Noisy Speech Signals for Hearing Aids

Abstract

a major part of the interaction between humans takes place via speech communication. It is very difficult to understand speech signals in presence of background noise for the normal listeners and hearing impaired persons. The human speech and hearing organ is inherently sensitive to interfering noise. Interfering noise decreases speech intelligibility and quality. Speech enhancement algorithms reduces the noise and improve one or more perceptual aspects of noisy speech most notably quality and intelligibility. This paper reports an importance of correct noise estimation in speech enhancement for hearing aids. Different environments are such as restaurant, train and Car are considered. Clean speech signals are corrupted by background noise respectively restaurant noise (multi-talker babble), train noise, and car engine noise at three different signal-to-noise ratio levels 0dB, 5dB, 10dB. Three different ways are used to estimate the noise. One is voice Activity Detector (VAD), second is Martins noise estimation algorithm, and third is MCRA-2 noise estimation algorithm. Modified spectral subtraction algorithm is used for speech enhancement. Subjective and objective type evaluation of enhanced speech signals were carried out and performance of VAD, Noise Estimation algorithms are evaluated. Section I introduction, section II explains basic Spectral subtraction and its modified version speech enhancement algorithm in detail and, section III explains different ways of noise estimation section IV explains performance evaluation of modified spectral subtraction algorithm along with various noise estimation techniques in terms of objective quality measures for Hearing aids, section-V conclusion.