Rayleigh modeling of teager energy operated perceptual wavelet packet coefficients for enhancing noisy speech

- A statistical model for determining an adaptive threshold is proposed using Rayleigh distribution function of the teager energy (TE) operated perceptual wavelet packet coefficients.
- A custom thresholding function is proposed that combines µ-law and semisoft thresholding functions and is able to provide better thresholding than that provided by the thresholding functions described in the literature.
- Detail simulations have been carried out in order to investigate the performance of the proposed method in terms of objective and subjective senses.
- The performance of our proposed method is compared with state-of-the-art methods, namely Universal and SMPO.
- Simulation results show that the proposed method yields consistently better results in the sense of higher output segmental SNR improvement in dB, higher output PESQ, and lower WSS values than those of the existing methods.

For enhancing noise corrupted speech, Rayleigh modeling of Teager energy (TE) operated perceptual wavelet packet (PWP) coefficients of the noisy speech is proposed in this paper. In order to obtain an enhanced speech, a threshold is derived and applied on the PWP coefficients by employing a custom thresholding function, which is designed based on a combination of μ-law and semisoft thresholding functions. The effectiveness of the proposed method is evaluated for car and multi-talker babble noise corrupted speech signals through performing extensive simulations using the NOIZEUS database. The proposed method is found to outperform some of the state-of-the-art speech enhancement methods not only at high but also at low levels of SNR in terms of standard objective measures and subjective evaluations including formal listening tests.