Unlock instant, AI-driven research and patent intelligence for your innovation.

Speech recognition system and method with decoupled wake-up phrase

A technology of speech recognition and phrases, used in speech recognition, speech analysis, instruments, etc.

Active Publication Date: 2020-09-18
BEIJING DIDI INFINITY TECH & DEV
View PDF1 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, there are many hurdles to overcome to improve the accuracy of speech recognition

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
  • Speech recognition system and method with decoupled wake-up phrase
  • Speech recognition system and method with decoupled wake-up phrase
  • Speech recognition system and method with decoupled wake-up phrase

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0017] In some approaches, speech recognition enables the recognition and translation of audio signals of spoken words and phrases into text by a machine such as a computer or server. The machine can use a speech decoder to identify the best matching text (eg, the phrase sequence with the highest probability score). This condition can be expressed as W=arg max P(W|X), which can be further transformed into W=arg max P(X|W)P(W). Here, an acoustic model can be applied to obtain P(X|W), speaking the sequence of phrases W=W 1 W 2 The probability of ... will get the data (sequence of feature vectors) X = X 1 x 2 ..., the audio signal X is given by some definite text W. Furthermore, speech models can be applied to obtain P(W), i.e. the sequence of phrases W = W that a human might actually say 1 W 2 ...The probability. W = W that maximizes P(X|W) and P(W) 1 W 2 ...sequence that yields the highest probability score, resulting in machine-recognized speech corresponding to th...

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

Systems and methods are provided for speech recognition. An example method may be implementable by a server. The method may comprise adding a key phrase into a dictionary comprising a plurality of dictionary phrases, and for each one or more of the dictionary phrases, obtaining a first probability that the dictionary phrase is after the key phrase in a phrase sequence. The key phrase and the dictionary phrase may each comprise one or more words. The first probability may be independent of the key phrase.

Description

technical field [0001] The present application relates generally to methods and techniques for speech recognition. Background technique [0002] Advances in technology allow users to use their voices for control. For example, commands for triggering various applications or functions input through a keyboard, mouse, or touch screen can be implemented through voice. However, there are many hurdles to overcome to improve the accuracy of speech recognition. Contents of the invention [0003] Various embodiments of the present application include systems, methods, and non-transitory computer-readable media for speech recognition. The exemplary method can be implemented by a system for speech recognition, such as a server. The method may include adding a key phrase to a dictionary comprising at least two dictionary phrases, and for each or more of the dictionary phrases, obtaining the first word of the dictionary phrase following the key phrase in the sequence of phrases. pr...

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): G10L15/06
CPCG10L15/187G10L15/197G10L15/02G10L15/18G10L15/183G10L15/22G10L15/26G10L15/30G10L2015/025G10L2015/088
Inventor 黄辰
Owner BEIJING DIDI INFINITY TECH & DEV