Estimation of speech embedded in a reverberant and noisy environment by independent component analysis and wavelets

摘要

In this paper, we develop a system for enhancement of the speech signal with highest energy from a linear convolutive mixture of n statistically independent sound sources recorded by m microphones, where m>n. In this system we use the concept of independent component analysis (ICA) along with adaptive auditory filter banks and pitch tracking. Computer simulations and real-world experiments carried out in an actual room and measured through objective and subjective measures confirm the validity of the proposed algorithm.