Speech enhancement method based on time-frequency domain joint loss function

Abstract

The invention provides a speech enhancement method based on a time-frequency domain joint loss function. The method comprises steps: integrating a clean voice data set and a noise data set in an open source data set into a noisy voice data set, converting the noisy voice data set into an amplitude spectrum, a phase spectrum and waveform data through preprocessing operation, and constructing a training set; constructing a CNN network model, taking the noisy voice amplitude spectrum as input, taking the clean voice amplitude spectrum as a label, and carrying out model training; performing waveform reconstruction on an amplitude spectrum estimation value output by the model and a noisy speech phase spectrum through an inverse short-time Fourier transform method to obtain a time domain waveform of estimated speech; calculating frequency domain loss through the clean voice amplitude spectrum and the amplitude spectrum estimated value; calculating time domain loss through the clean voice time domain waveform and the estimated voice time domain waveform; and constructing time-frequency domain joint loss according to the frequency domain loss and the time domain loss, and guiding the CNN network model to perform weight optimization. The phenomenon that the estimated amplitude spectrum is not matched with the phase spectrum is reduced, and the speech enhancement effect is improved.

Description

technical field [0001] The invention relates to the field of speech enhancement, in particular to a speech enhancement method based on a time-frequency domain joint loss function. Background technique [0002] Voice communication is the most convenient way of information interaction between people and machines. However, no matter what the environment is, voice communication is more or less disturbed by ambient noise. Speech enhancement technology is an effective way to solve the impact of noise in the process of voice interaction. The purpose of speech enhancement is to extract clean speech signals from background noise as much as possible, eliminate environmental noise, and improve speech quality and speech intelligibility. [0003] In recent years, the popularity of artificial intelligence technology has remained high, and speech enhancement technology has also developed rapidly, and various speech enhancement technologies emerge in endlessly. These speech enhancement s...

AI Technical Summary

Problems solved by technology

The mismatch between the magnitude spectrum and the phase spectrum will damage the characteristic information of the speech signal, such as the Mel Frequency Cepstrum Coefficient (MFCC), thus affecting the accuracy of speech recognition.

[0009] It can be seen that in the methods of the prior art, there are still defects in the speech enhancement method as the speech recognition front-end module, and it is of great practical significance to solve the problem of speech enhancement phase mismatch

Images