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

Hiaden Markov model edge decipher data reconstitution method f speech sound identification

A Hidden Markov and Data Reconstruction technology, used in speech recognition, speech analysis, instrumentation, etc.

Inactive Publication Date: 2004-02-18
INST OF ACOUSTICS CHINESE ACAD OF SCI
View PDF0 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

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
  • Hiaden Markov model edge decipher data reconstitution method f speech sound identification
  • Hiaden Markov model edge decipher data reconstitution method f speech sound identification
  • Hiaden Markov model edge decipher data reconstitution method f speech sound identification

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0036] The human ear's perception of sound has obvious nonlinear characteristics. Incorporating some factors that reflect the auditory characteristics of the human ear into the speech features can significantly improve the performance of the speech recognition system. Considering the critical band effect of the auditory system, the US frequency domain is usually used The triangular filter bank uniformly distributed on the upper part is used to analyze the subband characteristics of the speech feature vector, which has been widely used in speech recognition technology. In the following, the data reconstruction algorithm based on Hidden Markov Model Marginalized Viterbi decoding will be described by taking the data reconstruction of the subband eigenvector of Speech Beauty (Mel) as an example.

[0037] After missing component estimation, the speech feature S is divided into two vectors: the "missing vector" S m and "reliable vector" S o , figure 1 The missing component estimat...

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

A method for reconfiguring the marginalized decode data of hidden Markovian model (HMM) used in speed recognition features that the HMM transfer probability array is used to describe the dynamic characteristics of speech characteristic vector in time domain, the complete variance array is used to describe the relative characteristics between the components of the characteristic vector for Meizi band, and a data reconfiguring algorithm (VITDI) is used to reconfigured "lost vector". It can improve the noise robustness of speech recognition system.

Description

technical field [0001] The method of the invention relates to the application technology of computer technology, especially in the speech recognition technology, according to the speech feature not covered by the noise, the technology of estimating the speech feature damaged by the noise by using the marginalized Viterbi decoding process. Background technique [0002] Noise robustness is one of the main challenges that speech recognition technology is currently facing. In-depth research on the robustness of speech recognition based on data reconstruction has important theoretical significance and wide application value. [0003] When two sounds of unequal loudness act on the human ear, the presence of the louder frequency component will affect the perception of the lower loudness frequency component, making it less perceptible. This phenomenon is called the masking effect. According to the masking effect of the human ear, a data reconstruction method is proposed. The data r...

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
IPC IPC(8): G10L15/14G10L15/20
Inventor 杜利民罗宇
Owner INST OF ACOUSTICS CHINESE ACAD OF SCI
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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com