Multi-voice region speech recognition method, device and storage medium

Abstract

Embodiments of the present invention provide a multi-sound region speech recognition method, device, and computer-readable storage medium. Wherein the multi-sound range voice recognition method comprises: using a microphone module to collect voices in the voice collection environment, wherein, setting a target sound range for each microphone in the microphone module in advance, and the target sound range is preset One of a plurality of independent sound zones in the voice collection environment; judging whether the voice comes from a target sound zone; suppressing voices from other sound zones except the target sound zone; suppressing the voice to identify. The embodiment of the present invention uses a microphone module to isolate different sound zones, and by isolating multiple sound zones, all employees in the voice collection environment can use voice, and the isolation degree of each sound zone can reach 15-20dB, realizing ideal isolation Effect.

Description

technical field [0001] The present invention relates to the field of information technology, in particular to a multi-sound region speech recognition method, device and computer-readable storage medium. Background technique [0002] The traditional speech recognition technical solutions all focus on serving a target user. For example, the traditional in-vehicle voice recognition technology solution is designed around the driver alone, and the angle of the front-end microphone is aimed at the main driving position. The traditional core vertical navigation, music, phone, etc. are also designed around the driver. Among them, the vertical category is a vertical field, specifically referring to a certain field. Developers can set specific fields according to user usage scenarios to make the recognition results more accurate. For example, if the vertical field is set to "music", then the name of the song, artist name or lyrics related to music will be recognized preferentially d...

AI Technical Summary

Problems solved by technology

[0003] The defect of the traditional voice recognition technology solution is that it only serves one target user, and it is impossible to use voice services for all employees in the voice collection environment.

Still taking in-vehicle voice recognition as an example, the traditional voice recognition technology solution only serves the driver alone, and cannot realize the voice service for all people in the car, which brings poor user experience to the passengers in the car

Images