Noise removal in speech signal using fractional fourier transform

摘要

The Fractional Fourier Transform (FRFT) technique is derived from classical Fourier Transform. It is used in many applications. In this paper Fractional Fourier Transform is applied to denoise the noisy speech signal. The white Gaussian Noise is considered here which is affecting the speech signal. The performance of Fractional Fourier Transform is evaluated for different SNR and hard thresholding value. The results are presented by the measuring parameters MAE and PSNR. It is observed with fewer thresholds and SNR value, the MAE is minimum and PSNR is good.