Method for realizing background noise suppressing based on multiple statistics model and minimum mean square error

Abstract

The invention discloses an inhibiting method of background noise based on multiple statistical models and minimum mean square deviation, which comprises the following steps: proceeding short-time Fuliye transmission for voice signal corresponding to present input frame; utilizing the frame to retain the pure voice amplitude variance estimation and noise amplitude variance estimation on each frequency; calculating the estimation of real part and imaginary part corresponding to each frequent component of voice signal in the present frame; counting the probability of non-existed prior voice of each component frequency of the input frame; modifying the estimation result of real part and imaginary part of each frequency component; obtaining the true distribution of voice and noise in the real applying precisely; improving the inhibiting effect with stable estimating course; reducing the calculating complexity; fitting for kinds of voice communicating systems.

Description

technical field [0001] The invention belongs to the field of speech processing, and mainly relates to a background acoustic noise (Acoustic Noise) suppression method based on a multiple statistical model (Multiple Statistical Model) and a minimum mean square error (Minimum Mean Squared Error). Preprocessing for recognition and other applications. Background technique [0002] In most speech communication applications, the input end of the system can only receive noisy speech interfered by background noise. The noise greatly interferes with the quality of speech communication and reduces the clarity and intelligibility of speech. Voice processing modules such as encoding and decoding have adverse effects. [0003] The background noise suppression technology can extract the original speech as pure as possible from the noisy speech. This research belongs to the category of "speech enhancement" in the field of speech processing. Noise suppression helps to improve the perceived...

AI Technical Summary

Problems solved by technology

[0012] The key issue of the MMSE method is which statistical model to use to reflect the statistical distribution of the frequency-domain speech signal. The Gaussian model is commonly used in the existing methods, but it cannot simulate the real situation well; on the other hand, the accuracy of the noise estimation Robustness and robustness are also key factors in determining the performance of suppression methods

Images