A method for automatic adjustment of speech volume based on energy statistics

Abstract

The invention discloses a voice volume automatic adjustment method based on energy statistics. The present invention includes the following steps: (1) Counting the energy average value energy of each frame sampling point and the maximum peak peak peak in the absolute value of the energy values ​​of all sampling points in the frame; (2) calculating the sampling of each frame in the frame from 0 to frame_index The average value of the energy average energy of the point energy_avg and the average value of the maximum peak value peak_avg; (3) Calculate the enlargement factor factor_max' and the reduction factor factor_min' in the next time period time; (4) For the next time period time Each frame of voice data in the frame is determined, and when zooming is required, the zoom factor is used to zoom in or out; (5) The processed voice data frame is output, and the end is ended. The invention utilizes the similarity and continuity of speech data to predict the amplification factor and reduction factor to be used in the next period of time according to the statistical information of the speech data in the previous period of time, and reduces the complexity of the algorithm as much as possible while satisfying the practicability.

Description

technical field [0001] The invention belongs to the field of computer digital voice processing and communication, in particular to an automatic voice volume adjustment method based on energy statistics. Background technique [0002] In the field of voice processing and communication, such as online education systems, video conferencing systems, etc., the voice data input from the microphone may be too small or too large due to the influence of various situations. The voice volume is adjusted accordingly through the volume adjustment module. Increase or decrease can maintain a relatively stable volume level, which makes people sound more comfortable. [0003] In various speech processing communication systems, most of the systems have the function of automatic volume adjustment. There are two main implementation methods. One is to change the volume of the source voice collected from the microphone by adjusting the volume of the system microphone. It is realized by enlarging ...

AI Technical Summary

Problems solved by technology

When it is realized by adjusting the volume of the system microphone, when it is judged that the volume needs to be increased or decreased, by calling the microphone interface of the system, the gain and volume of the microphone are increased and decreased accordingly to achieve automatic volume adjustment. The advantage is that the software can be reduced. The amount of calculation required for processing will not affect the voice quality. The disadvantage is that the system volume will be adjusted frequently, which will affect the user experience; when processing in the time domain through software, the pcm voice data will be directly scaled The advantages of calculation are that the algorithm is simple and the amount of calculation is small. The disadvantage is that in theory, the noise in some voices will also be enlarged and reduced accordingly, but it is found that the impact on the user experience is not great during actual use. When processing, it is necessary to transform the pcm data into the frequency domain first, and then perform corresponding processing in the frequency domain, and then convert the data in the frequency domain into pcm data through inverse transformation. The advantage of processing in the frequency domain is that it can be used for each frequency band Data is controlled, and the required frequency band and data are scaled more purposefully. The disadvantage is that the algorithm is complex and the amount of calculation is relatively large.

When adjusting the volume of voice data through software, the advantage is that there is no need to operate the volume of the system microphone, which will not affect the system volume. The disadvantage is that it requires a certain amount of calculation and will have some impact on the sound quality.

Images