Speech recognition method and system based on AC automaton hot word enhancement

Abstract

The invention provides a speech recognition method and system based on AC automaton hot word enhancement, and belongs to the technical field of speech recognition. Respectively constructing language models taking characters and words as units for generating a decoding graph; a hot word list is obtained, words and words in the language model serve as a dictionary, word segmentation is conducted on each hot word in the hot word list, and an AC automaton is constructed; the method comprises the following steps: constructing a token structure body in a decoding path search process of speech recognition, wherein the token structure body is used for storing all node pointers and total cost of a current decoding path on a decoded graph, a node pointer of a current matching path on an AC automatic machine and a reduction cost value of a node; the token structural body is transmitted along with a node decoded at the current moment in the decoded graph; and dynamically matching the recognition result output by the current node in the decoded image with the AC automaton, establishing a cache mechanism in the dynamic matching process, updating the total cost value according to the matching result, taking the path with the minimum cost as a target path, and outputting a speech recognition result.

Description

technical field [0001] The invention relates to the technical field of voice recognition, in particular to a voice recognition method and system based on AC automaton hot word enhancement. Background technique [0002] Speech recognition technology is often applied in various specific fields in the production environment. Each field has its specific proper nouns, and it is difficult for a speech recognition system in general scenarios to accurately recognize these proper nouns. Hot word enhancement refers to improving the recognition rate of hot words in speech recognition results based on proper noun hot words provided by users. [0003] In the prior art, the traditional speech recognition system solves the hot word recognition mainly by integrating the hot words into the language model. This method cannot modify the hot word list. Once modified, it is necessary to rebuild and set the training language Another way is to match the recognition results with the hot words in ...

AI Technical Summary

Problems solved by technology

[0003] In the prior art, the traditional speech recognition system solves the hot word recognition mainly by integrating the hot words into the language model. This method cannot modify the hot word list. Once modified, it is necessary to rebuild and set the training language Another way is to match the recognition results with the hot words in the hot word list one by one. The matching time is long and the efficiency is low, and the increase of hot words will double the recognition speed, and the effect of hot word enhancement is very limited.

Images