A blind source separation with fractional calculus for noise reduction in speech enhancement

2017 International conference on Microelectronic Devices, Circuits and Systems (ICMDCS) Pub Date : 2017-08-01 DOI:10.1109/ICMDCS.2017.8211701

Pramodini Shrihari Talware, V. Tank, S. Mahajan

{"title":"A blind source separation with fractional calculus for noise reduction in speech enhancement","authors":"Pramodini Shrihari Talware, V. Tank, S. Mahajan","doi":"10.1109/ICMDCS.2017.8211701","DOIUrl":null,"url":null,"abstract":"A Blind Source Separation with Fractional Calculus for Noise Reduction in Speech enhancement is proposed in this paper. Different strategies are available for reduction of noise. Fractional calculus has been recently applied in various zones like engineering, science, bio-engineering and finance. It has numerous applications, for example, use in differentiation, integral equations, signal processing, fluid mechanics, and electrochemistry. In this work speech processing signal application where Discrete Fractional Fourier Transform (DFRFT) is used which is an essential process for signal processing. DFRFT has DFT hermite Eigenvectors and retains the eigenvalue eigen operate relation as a fractional fourier transform that reconstruct the signal. For the purpose of noise reduction, Blind Source Separation has been utilizing which does not have prior knowledge of original signal. DFRFT algorithm and SNR are used to prove that the improvement of the processed enhanced signal as compared to the noisy signal.","PeriodicalId":314717,"journal":{"name":"2017 International conference on Microelectronic Devices, Circuits and Systems (ICMDCS)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2017-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2017 International conference on Microelectronic Devices, Circuits and Systems (ICMDCS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICMDCS.2017.8211701","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 1

Abstract

A Blind Source Separation with Fractional Calculus for Noise Reduction in Speech enhancement is proposed in this paper. Different strategies are available for reduction of noise. Fractional calculus has been recently applied in various zones like engineering, science, bio-engineering and finance. It has numerous applications, for example, use in differentiation, integral equations, signal processing, fluid mechanics, and electrochemistry. In this work speech processing signal application where Discrete Fractional Fourier Transform (DFRFT) is used which is an essential process for signal processing. DFRFT has DFT hermite Eigenvectors and retains the eigenvalue eigen operate relation as a fractional fourier transform that reconstruct the signal. For the purpose of noise reduction, Blind Source Separation has been utilizing which does not have prior knowledge of original signal. DFRFT algorithm and SNR are used to prove that the improvement of the processed enhanced signal as compared to the noisy signal.

查看原文本刊更多论文

基于分数阶演算的语音增强降噪盲源分离

提出了一种基于分数阶演算的盲源分离方法用于语音增强降噪。减少噪音有不同的策略。近年来，分数阶微积分在工程、科学、生物工程和金融等领域得到了广泛的应用。它有许多应用，例如，用于微分，积分方程，信号处理，流体力学和电化学。在语音处理信号的应用中，离散分数阶傅里叶变换(DFRFT)是信号处理的一个重要过程。DFRFT具有DFT埃尔米特特征向量，并保留特征值与特征运算关系作为重构信号的分数阶傅里叶变换。为了达到降噪的目的，采用了不知道原始信号先验信息的盲源分离方法。用DFRFT算法和信噪比证明了处理后增强信号相对于含噪信号的改善。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

2017 International conference on Microelectronic Devices, Circuits and Systems (ICMDCS)

自引率

0.00%

发文量