A short-term speech duration extension method for language recognition

A language recognition and speech technology, applied in speech recognition, speech analysis, natural language data processing, etc., can solve problems such as poor language recognition performance, and achieve the effect of reducing interference

Active Publication Date: 2020-03-17
INST OF ACOUSTICS CHINESE ACAD OF SCI +1
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] The purpose of the present invention is to overcome the problem of poor language recognition performance of current short-duration speech, and propose a short-duration speech duration extension method applied to language recognition, which uses speech time domain stretching technology to directly extend the duration of the speech to be recognized ; For each speech to be recognized, after generating multiple speeches at different speech rates, they are spliced ​​with the original speech to form a longer speech

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
  • A short-term speech duration extension method for language recognition
  • A short-term speech duration extension method for language recognition
  • A short-term speech duration extension method for language recognition

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0033] The present invention will be further described now in conjunction with accompanying drawing.

[0034] like figure 1 Shown, a kind of short-term speech duration extension method applied to language recognition, said method comprises:

[0035] Step 1) For a speech x to be recognized, its duration is length(x), judge whether the length(x) is less than the threshold T, if the judgment result is affirmative, go to step 2), otherwise, do not need to process the speech ;

[0036] Step 2), determine the quantity n of the different speech rate voices that generate; N determines according to the duration of input voice:

[0037]

[0038]It can be seen from the calculation formula of n that the shorter the input voice duration, the more voices need to be generated.

[0039] Step 3), fix the composite frame shift to S s , according to the speech rate change rate, select n decomposition frame shift S a value of:

[0040] The speech rate change rate α is defined as:

[0...

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 provides a short-duration voice duration extension method applied to language recognition. The method comprises the following steps: for to-be-recognized voice with short duration, determining the number n of generated voices with different voice speeds according to the duration of the voice; carrying out calculating according to a synthesis frame shift value and n voice speed change rates, so as to generate n decomposition frame shift values of the voice; and generating n voices with different voice speeds according to decomposition frame shift values and the synthesis frame shift value, and splicing the n voices with different voice speeds and the original voice, so as to generate the voice with the extended duration. The language information of the voices with different voice speeds is complementary, and the method provided by the invention can remarkably promote the language recognition performance of the short-duration voice.

Description

technical field [0001] The invention relates to the field of computer language recognition, in particular to a short-term speech duration extension method applied to language recognition. Background technique [0002] Language recognition refers to the technology that a computer automatically determines the language category of a speech. This is a technology that enables large-scale cross-lingual speech recognition applications, such as spoken language translation, spoken document retrieval, and more. It is also a research hotspot in information extraction in the field of intelligence and security. [0003] The speech duration to be recognized is too short, which is a common problem in research fields such as speaker recognition and language recognition. In recent years, there have been some targeted studies on short-term speech recognition. Reference [1] (A.K.Sarkar, D.Matrouf, P.Bousquet, and J.Bonastre. Study of the effect of i-vector modeling on short and mismatch utt...

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): G10L15/00G10L15/04G10L15/10G06F40/263
CPCG06F40/205G06F40/263G10L15/005G10L15/04G10L15/10
Inventor 周若华袁庆升张健颜永红包秀国
Owner INST OF ACOUSTICS CHINESE ACAD OF SCI
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

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap