Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Speech recognition method and wake-up word detection method and device

A technology of speech recognition and speech recognition model, applied in speech recognition, speech analysis, instruments, etc., can solve the problems of poor detection performance and word-repetition wake-up word design, achieve low false wake-up rate, improve detection performance, and high wake-up rate effect

Active Publication Date: 2019-01-11
TENCENT TECH (SHENZHEN) CO LTD
View PDF10 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In the context of Chinese language applications, choosing redundancies as wake-up words is very popular among users, and choosing redundancies-type wake-up words (such as "Tencent Tencent") has good wake-up performance. However, based on the currently provided wake-up word detection methods, Reasonable schemes have not been designed for redundant wake words, so in voice interactive devices, the detection performance of redundant wake words is poor

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 method and wake-up word detection method and device
  • Speech recognition method and wake-up word detection method and device
  • Speech recognition method and wake-up word detection method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0047] Embodiments of the present invention provide a method for speech recognition, a method and a device for detecting wake-up words, which fully consider the pronunciation characteristics of redundancies, and use the speech recognition model to calculate the score of each word in redundancies. Only when the score meets the conditions can it be recognized as a wake-up word, thereby improving the performance of detecting redundancies-style wake-up words, with a higher wake-up rate and a lower false wake-up rate.

[0048] The terms "first", "second", "third", "fourth", etc. (if any) in the description and claims of the present invention and the above drawings are used to distinguish similar objects, and not necessarily Used to describe a specific sequence or sequence. It is to be understood that the data so used are interchangeable under appropriate circumstances such that the embodiments of the invention described herein are, for example, capable of practice in sequences othe...

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 invention discloses a wake-up word detection method, which comprises the following steps: obtaining speech information to be recognized, wherein the speech information to be recognized comprises atarget word set, the target word set is composed of at least two groups of repetitive target words, and each group of target words comprises at least one target word; acquiring a target speech feature vector according to the speech information to be recognized; using the speech recognition model for outputting the probability scores corresponding to each word in the speech information, wherein the target probability scores set includes the target probability scores corresponding to each target word. If the target probability score set satisfies the speech awakening condition, it is determinedthat the speech information to be recognized includes a target awakening word, and the target awakening word belongs to one awakening word in the awakening word set. The invention discloses a wake-upword detecting device and a terminal device. The invention improves the detection performance of the reduplicated awakening word, and has high awakening rate and low false awakening rate.

Description

technical field [0001] The invention relates to the field of artificial intelligence computers, in particular to a voice recognition method, a wake word detection method and a device. Background technique [0002] With the popularization of smart voice devices, human-computer interaction with smart devices through voice commands has become an important function. In voice interaction applications, users can wake up smart devices through voice wake words, so the wake word detection performance of smart devices has a significant impact on the human-computer interaction experience. [0003] At present, a method based on dynamic time warping (DTW) can be used for wake-up word detection. This method dynamically matches the acoustic features of the voice with several pre-recorded recordings of the same wake-up word, and compares the matching distance with the pre-recorded Compared with the set threshold, when the distance is less than the threshold, it is judged that the voice is ...

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 Applications(China)
IPC IPC(8): G10L15/22G10L15/197G10L15/02
CPCG10L15/02G10L15/197G10L15/22G10L2015/223
Inventor 高毅陈杰苏丹于蒙罗敏
Owner TENCENT TECH (SHENZHEN) CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products