Speech recognition method, device and electronic equipment

Abstract

Provided are a method and device for speech recognition, and an electronic device. The method for speech recognition comprises: acquiring speech information to be recognized (S110); recognizing the speech information to be recognized, acquiring current speech recognition information corresponding to the speech information to be recognized (S120); determining whether stored speech recognition information to be spliced is present (S130); if present, splicing the speech recognition information to be spliced with the current speech recognition information, producing spliced speech recognition information (S140); determining whether the spliced speech recognition information has complete semantics (S150); if yes, then determining the spliced speech recognition information as a speech recognition result (S160); and if not, determining and storing the spliced speech recognition information as speech recognition information to be spliced, and continuing executing the step for acquiring speech information to be recognized (S170). The method for speech recognition ensures the completeness of the semantics recognized and increases speech recognition effects for discontinuous speech.

Description

technical field [0001] The invention relates to the technical field of voice recognition, in particular to a voice recognition method, device and electronic equipment. Background technique [0002] Currently, many smart devices have voice recognition functions. Generally speaking, the function of speech recognition can be realized in the following two ways: [0003] One is: the smart device receives the voice command information, recognizes the voice command information, obtains the recognized command information, and responds to the recognized command information. [0004] The other is: the smart device receives the voice command information, and sends the voice execution information to the cloud server, and the cloud server recognizes the voice command information, obtains the recognized command information, and responds to the recognized command information, Return the response information to the smart device. [0005] In real life, when users speak voice command infor...

AI Technical Summary

Problems solved by technology

[0007] That is to say, due to the existence of silent segments in incoherent speech, when the existing speech recognition methods recognize such incoherent speech, the recognized semantics are usually incomplete, which affects the speech recognition effect.

Images