Voice excitation control method and device

Abstract

The embodiment of the invention discloses a voice excitation control method, wherein the method comprises the steps: sampling the voice of each client in a preset sampling period at a preset samplingrate, and taking the mean value of the volumes of all sampling points in a sampling period as the volume reference value of the corresponding client in the sampling period; taking the client corresponding to the maximum value of the volume reference values in the sampling period as a volume representation terminal of the sampling period; taking a continuous threshold number of sampling periods asan acquisition period, and calculating the number of times that each client in the acquisition period serves as the volume representative terminal; and taking the client with the highest frequency asa current voice excitation terminal. According to the method, it is guaranteed that a speaker cannot be frequently switched due to sudden sound, it is guaranteed that in the scene of a network conference, the speaker and a main screen image are kept consistent, and the purpose of consistent audio signals is achieved.

Description

technical field [0001] The invention relates to the field of video conferencing, in particular to a voice activation control method and device. Background technique [0002] With the development of telecommunication technology, the video conferencing service has been more and more widely used. In the conference video system, it is necessary to determine the multiple speakers in the conference and mix their voices to make the participants feel more natural; at the same time, the image of the speaker must be broadcast to other participants. The conference video system takes the video conference multi-point control unit (MCU) as the core, responsible for image switching and sound mixing processing of all conference terminals. Since there are usually multiple participants in the conference TV, it is necessary to control and manage the entire conference, that is, conference control, such as switching the speaker site, controlling the audio mixing, and the speaker selects the sit...

AI Technical Summary

Problems solved by technology

At present, the problems existing in the conference scene include: no voice activation is implemented, switching the viewing angle or locking a plane through the wheel screen, the interactivity will be relatively poor, and it is difficult to focus on the speech when someone is speaking; the realization of voice activation Based on theory, it has not been actually applied to the product; the voice activation algorithm or solution is not optimized enough, resulting in frequent switching of the main screen, which cannot be locked, and reduces the customer experience

Images