2. Speech enhancement based on spectral subtraction and phase compensation
Utilizing the benefits of spectral subtraction and phase compensation,
I developed some methods, that can enahnce speech in adverse nevironments such as
at -10 dB SNR.
1. Speech enahncement based on statistical modeling of speech signal:
I developed several thresholding based speech enahancement methods in perceptual wavelet packet domain
that perform better than other available methods.