Method and system for improving remote field voice recognition rate, and readable storage medium

Abstract

The invention provides a method and system for improving the remote field voice recognition rate, and a storage medium. An intelligent television scans the front side of an intelligent television in real time to obtain the position, the motion and feature information of the front object; the position of a sound source giving out an activation word is detected and positioned through a wave arrivaldirection positioning technology; whether the sound source is people or not is judged through feature information obtained through radar; if the sound source is people, a narrow sound picking wave beam pointing to the sound source direction is formed through an MIC array for sound picking; the sound source is positioned in real time through radar; the sound picking wave bead regulates the angle totrack the sound source in real time according to the position parameter of the sound source provided by the radar; the sound picking stability is improved. Compared with the prior art with the defects of difficult distinguishing between an interference source and a target sound source, unprecise positioning and sound picking incapability during the target sound source movement, the method and thesystem provided by the invention have the advantages that the precise positioning can be realized; the interference source and the target sound source can be distinguished; the target sound source can be tracked; the remote field voice recognition rate is improved.

Description

technical field [0001] The invention relates to a smart TV technology, in particular to a method, a system and a readable storage medium for improving the far-field voice recognition rate. Background technique [0002] With the development of smart TV technology, AI has gradually become the standard configuration of TV products, and the realization of voice interaction has become the development direction of AI smart TV. Many AI smart TVs currently on the market support voice interaction. The so-called far-field voice, as the name suggests, means that smart devices such as AI smart TVs can also hear and understand the user's voice commands in a relatively distant place (such as 5 meters away). It is a change that combines AI voice search. The innovative interactive function of the experience, in the far-field voice interaction, how to easily wake up and control the smart TV device with voice in the complex living room environment has become an extremely important user exper...

AI Technical Summary

Problems solved by technology

[0004] In view of the shortcomings of the above-mentioned prior art, the present invention proposes a method for improving the far-field speech recognition rate in order to solve the defects and deficiencies of the prior art, by integrating the radar positioning technology and the direction of arrival positioning technology on the smart TV (DOA), collect, process and judge the information of each sound source, accurately locate and distinguish the interference source and the target sound source, combined with the built-in far-field voice beamforming function, form a sound pickup beam with a small angle, only pointing to the target sound Source and pick up the sound of the target sound source to achieve the purpose of accurate sound pickup. Due to the real-time nature of radar positioning, the angle of the beam can be adjusted in real time to track and pick up the sound of the target sound source, which solves the problem that the target sound source cannot be picked up when the position of the target sound source moves. sound problem

Images