Improved self-coding neural network voice enhancement algorithm

Abstract

The invention discloses an improved self-coding neural network voice enhancement algorithm and aims at problems in the voice enhancement effect existing in traditional voice enhancement algorithms such as a spectral subtraction algorithm and a Wiener filtering method, e.g., the poor filtering effect for the non-stable noise, residual of music noise after enhancement and the poor generalization effect for noise types and the signal to noise ratio. According to the algorithm, a three-level junction of the self-coding neural network is enhanced to be a five-level junction, the neuron numbers of the correspondingly levels are respectively 256, 128, 64, 128 and 256, 100 types of noise are added to pure audios according to signal to noise ratios of -5dB, 0dB, 5dB, 10dB and 15dB, mass data sets are constructed to train the network, the excellent voice enhancement effect can be realized after the network model is trained, as the training data is big, the excellent generalization effect for thenoise types and the signal to noise ratio is realized.

Description

technical field [0001] The invention relates to the technical field of speech signal processing, in particular to an improved self-encoding neural network speech enhancement algorithm. Background technique [0002] In recent years, speech recognition technology has developed rapidly, and the pursuit of higher recognition rate has also accelerated the rapid development of speech enhancement technology. Looking at the development of speech enhancement for decades, great progress has been made in theory and practice. Time The enhancement algorithms in the frequency domain include parameter and filtering methods, and the more mature algorithms in the frequency domain include spectral subtraction and Wiener filtering. These mature algorithms can indeed achieve good enhancement effects on stationary noise. [0003] However, both time domain and frequency domain enhancement algorithms have the following disadvantages. First of all, these algorithms will add some unreasonable assump...

AI Technical Summary

Problems solved by technology

[0003] However, both time domain and frequency domain enhancement algorithms have the following disadvantages. First of all, these algorithms will add some unreasonable assumptions for the convenience of pushing to the process, and these unreasonable assumptions will definitely affect Speech enhancement effect. Secondly, these enhancement algorithms can have a good filtering effect on stationary noise signals, but the filtering effect on non-stationary noise signals is not good enough. Moreover, there will still be a small amount of noise remaining in the audio after speech enhancement. "Music noise" will also be added. Finally, there are many types of noise in real life, and the range of signal-to-noise ratio is large.

Images