Voice stream processing method and device

A technology of speech stream and processing results, applied in speech analysis, speech recognition, instruments and other directions, can solve the problems of large memory, slow recognition speed, long time consumption, etc., to achieve the effect of less search paths, fast recognition speed, and improved processing speed

Active Publication Date: 2022-06-07
北京葡萄智学科技有限公司
View PDF9 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Correspondingly, in the online decoding stage, the acoustic model, language model, and pronunciation dictionary need to be built into a large decoding resource. The memory occupied by the decoding resource is very large, ranging from several G to tens of G, and it is often necessary to load the decoding resource in advance. , takes a long time, and uses a very large decoding resource to identify the voice stream, and the recognition speed is slow

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Voice stream processing method and device
  • Voice stream processing method and device
  • Voice stream processing method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0076] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0077] In the process of realizing the present application, the inventor of the present application found that the voice stream processing method in the related art is: decoding the voice stream; Feature extraction is performed. In a large decoding resource, an optimal path is searched for the feature sequence of the speech stream as the recognition result of this decoding.

[0078] Among them, the specific decoding process is as follows: first...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The embodiment of the present application provides a voice stream processing method and device. The method includes: acquiring the voice stream to be recognized for the target task, and determining the target identifier of the target task; according to the target identifier, decoding small resources Search the group for a target decoding small resource that matches the target identifier; if the target decoding small resource is found, use the target decoding small resource to process the voice stream to be recognized, and obtain a processing result. Through this method, the decoding speed can be improved, and the memory occupied by decoding resources can be reduced.

Description

technical field [0001] The embodiments of the present application relate to the technical field of speech recognition, and in particular, to a method and apparatus for processing a speech stream. Background technique [0002] In the era of artificial intelligence, voice, as one of the entrances of human-computer interaction, is used more and more widely. The process of voice interaction between people and mobile phones, computers or other smart devices is as follows: first, the device collects the voice spoken by the person, then the device converts the collected voice into text, then parses the text, and finally the device gives the corresponding voice. instruction. In this interactive process, the technology of converting speech into text is called speech recognition; among them, Chinese speech recognition converts Chinese speech into Chinese character strings, and English speech recognition converts English speech into English word strings. [0003] In the related art, ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Patents(China)
IPC IPC(8): G10L15/02G10L15/06G10L15/26
CPCG10L15/26G10L15/063G10L15/02G10L2015/025
Inventor 王永庆张俊博佟子健
Owner 北京葡萄智学科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products