Evaluation of a new approach for speech enhancement algorithms in hearing aids

Abstract

Several studies on hearing impaired people who use hearing aid reveal that speech enhancement algorithms implemented in hearing aids improve listening comfort. However, these algorithms do not improve speech intelligibility too much and in many cases they decrease the speech intelligibility, both in hearing-impaired and in normally hearing people. In fact, current approaches for development of the speech enhancement algorithms (e.g. minimum mean square error (MMSE)) are not optimal for intelligibility improvement. Some recent studies investigated the effect of different distortions on the enhanced speech and realized that by controlling the amplification distortion, the intelligibility improves dramatically. In this paper, we examined, subjectively and objectively, the effects of amplification distortion on the speech enhanced by two algorithms in three background noises at different SNR levels.

Performance Improvement of Digital Hearing Aid Systems

“…It has been shown in [23] that, speech enhancement algorithms implemented in hearing aids improve listening comfort but they are not optimal for intelligibility improvement. Besides, it is shown in [23] that, when the amplification distortion is controlled, the intelligibility improves significantly.…”

“…Besides, it is shown in [23] that, when the amplification distortion is controlled, the intelligibility improves significantly. This parameter is employed to estimate the amount of attenuation and amplification distortions in the enhanced speech to obtain more effective speech enhancement algorithm with better performance in terms of quality and intelligibility.…”

“…This parameter is employed to estimate the amount of attenuation and amplification distortions in the enhanced speech to obtain more effective speech enhancement algorithm with better performance in terms of quality and intelligibility. The is expressed as [23]: (14) where is the true instantaneous SNR at frequency bin , is the enhanced SNR and and are wanted signal magnitude spectrum, the noise signal magnitude spectrum, and the estimated magnitude spectrum of the wanted signal respectively.…”

“…Also, [10] proposes a discriminative post-filter (DPF) approach that uses a non-linear function to further enhance the SNR and quality of GMAPA speech enhancement algorithm. Moreover, [13] evaluates the effects of amplification distortion on the speech enhanced by Logmmse and Spectral Subtraction algorithms and shows that in order to improve the intelligibility when speech enhancement algorithm is employed, the amplification distortion should to controlled. This paper focusses on noise reduction aspect of digital hearing aid systems in order to suppress noise that maintains a relative constant sound pressure level (SPL) over time.…”

“…In our previous work in [7] and [8], adaptive-control factor and multiband based spectral subtraction methods are implemented to enhance the performance of communication systems and to address the issue of the associated musical noise generated by the conventional spectral subtraction algorithm. In this work, the signal to residual spectrum ratio (SR) presented in [13] is implemented with multi-band based spectral subtraction methods in order to control the amplification distortion for speech intelligibility improvement. Also, noise-estimation algorithm proposed in [12] for highly non-stationary noisy environments is employed.…”