Publication: Noise reduction approach for Arabic phonemes articulated by Malay speakers
No Thumbnail Available
Date
2018
Authors
Almisreb A.A.
Tahir N.M.
Abidin A.F.
Din N.M.
Journal Title
Journal ISSN
Volume Title
Publisher
Institute of Electrical and Electronics Engineers Inc.
Abstract
Noise reduction in speech signal is significant due to its worthy applications like VoIP, Speech Recognition and Cellular calls. In this paper, a method for noise reduction in Arabic phonemes speech waveforms based on corpus collected from nonnative Arabic speakers is investigate, specifically Malay. Firstly, each speaker is required to articulate the 28 Arabic alphabets without any pre-training. The proposed method combines both Gamma correction function and Negative function, in order to eliminate and reduce the unvoiced signals. In addition, a modified equation of the Gamma correction function is proposed to be used separately from the negative function. Results attained showed that the proposed method capable to enhance the quality and the computational time as well. Based on the spectrogram, the proposed method has shown advantages as compared to Zero-Crossing Rate approach. Experimental results showed that reduction in noise is attained by increasing the gamma value. � 2018 IEEE.
Description
Speech recognition; Arabic. Malay; Computational time; Gamma correction; Gamma function; Modified equation; Speech segmentation; Speech waveforms; Zero crossing rate; Signal processing