A DCT-based noisy speech enhancement method using Teager energy operator

This paper introduces an improved Discrete Cosine Transform (DCT) based speech enhancement method using the Teager energy operator (TEO). In this method, the TEO is employed upon the DCT coefficients of the noisy speech in order to develop a time-adaptive threshold value with lower computational burden. Unlike conventional thresholding-based speech enhancement schemes, the proposed method improves the perceptual quality of the enhanced speech substantially by overcoming the the problem of over thresholding of speech segments. In addition, the proposed method does not require a complicated estimation of the noise level or any knowledge of the Signal to Noise Ratio (SNR). The experimental results show that the proposed method is significantly more effective in reduction of not only the white noise but also the color noise from the noisy speech signal with different levels of SNR, thus garnering better performance in standard objective measures as well as subjective evaluations compared to some recent state-of-the-art methods of noisy speech enhancement.