No abstract

No abstract

Aiming at the problem of auditory negative enhancement of typical phase reconstruction method, an improved method of phase reconstruction and MMSE-LSA estimation is proposed. First, the geometric relationship between noisy speech and clean speech in unvoiced segment is used to estimate the phase of the clean speech; Second, considering the randomness of speech appearance in the actual noise environment, a modified MMSE-LSA amplitude estimation is proposed by using the binary hypothesis model. Finally, the new phase reconstruction in voiced and unvoiced speech is combined with the modified MMSE-LSA. The simulation results show that the performance of the algorithm proposed in this paper is better than typical phase reconstruction method in terms of the SegSNR and PESQ.

A new algorithm based on information maximization is proposed, against the shortcomings of traditional speech blind source separation of low convergence rate and high crosstalk error. It uses the new sgn function to make mutual information of input and output to maximize by analyzing a variety of non-linear function of the separation performance, and advances based sgn function BSS with fixed step-size and adaptive variable step-size. Experiments show that the new algorithm has advantages, such as fast convergence, small crosstalk error and good separation efficiency which compared with traditional methods of Be11’s and Sejnowskl’s.

No abstract