Speech recognition method and device

Abstract

The invention provides a speech recognition method and a speech recognition device, applied to intelligent equipment. The method comprises the following steps: collecting speech data of a user, extracting voiceprint information in the speech data of the user, and training all extracted voiceprint information by virtue of a voiceprint trainer, so that speech characteristic information and a voiceprint model of the user; and when a speech command of the user is received, converting the voice command in accordance with the speech characteristic information and the voiceprint model of the user, sothat a voice command which conforms to a user's command format requirement of the intelligent equipment. According to the speech recognition method and device, real expression of the user can be extracted in combination with speech characteristics, and an accuracy rate of speech recognition can be improved.

Description

technical field [0001] The invention relates to the technical field of voice recognition, in particular to a voice recognition method and device. Background technique [0002] As the fastest and most efficient way for human communication, voice has not been well used in the field of machines. With the development of artificial intelligence technology, the advantages of voice in human-computer interaction have also been reflected, and corresponding products have emerged , such as voice assistant software for mobile terminals, smart speaker equipment, etc. [0003] During the use of voice assistants, recognition errors are often caused by the influence of external noise. It will also be difficult to recognize the real intention of the user due to the difference in personal expression. Existing voice assistants have processes to improve the recognition rate, such as noise reduction for user voice commands, fixed user expressions, and allowing users to customize their own comm...

AI Technical Summary

Problems solved by technology

[0004] At present, the mainstream noise reduction is realized by adding hardware, which is too costly, and it is difficult to achieve noise reduction purely through algorithms in software, and the user’s command format is fixed, which makes the user experience very unfriendly and less flexible. Low

Moreover, the voice assistants of smart devices or smart speakers do not collect all the voice data of users. They only conduct artificial intelligence training on the voice data collected after waking up. Due to the limitations of voice assistants on user instructions, many users' own voiceprints Features will be deliberately hidden, for example, users will fix their own expressions, deliberately speak Mandarin clearly, etc. As a result, these data cannot reflect the state of the user's daily life, making it difficult to extract the user's own voice features , the trained model also lacks pertinence, so the recognition rate of the user's voice is not ideal

Images