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

Sparse representation features for speech recognition

一种语音识别、语音的技术，应用在语音识别、语音分析、仪器等方向，能够解决无法先验地获知类别或类型等问题，达到提高语音识别性能的效果

Inactive Publication Date: 2012-04-18
INT BUSINESS MASCH CORP
View PDF5 Cites 12 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

It should be understood that although speech classification can be performed in speech recognition tasks, in this case the class or type is often not known a priori

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
  • Sparse representation features for speech recognition
  • Sparse representation features for speech recognition
  • Sparse representation features for speech recognition

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0012] The principles of the invention will be described herein in the context of exemplary embodiments of methods, apparatus, articles of manufacture and systems for providing speech recognition functionality. It should be understood, however, that the principles of the present invention are not limited to the particular methods, apparatus, articles of manufacture and systems exemplarily shown and described herein. Rather, the principles of the present invention relate broadly to speech recognition techniques by which recognition performance can be improved by generating and using sparsely represented features in an example-based training approach. To this end, numerous modifications may be made to the illustrated embodiment within the scope of the invention. That is, no limitation to the specific embodiments described herein is intended or should be inferred.

[0013] It has been recognized that the failure of existing example-based approaches to the recognition task can be...

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

Techniques are disclosed for generating and using sparse representation features to improve speech recognition performance. In particular, principles of the invention provide sparse representation exemplar-based recognition techniques. For example, a method comprises the following steps. A test vector and a training data set associated with a speech recognition system are obtained. A subset of the training data set is selected. The test vector is mapped with the selected subset of the training data set as a linear combination that is weighted by a sparseness constraint such that a new test feature set is formed wherein the training data set is moved more closely to the test vector subject to the sparseness constraint. An acoustic model is trained on the new test feature set. The acoustic model trained on the new test feature set may be used to decode user speech input to the speech recognition system.

Description

technical field [0001] The present invention relates generally to speech recognition, and more particularly to techniques for generating and using sparsely represented features to improve speech recognition performance. Background technique [0002] As we all know, Gaussian Mixture Model (GMM) has been very widely used in speech recognition problems. Although GMMs allow for fast model training and scoring, training samples are pooled together for parameter estimation, resulting in loss of information present in individual training samples. [0003] On the other hand, example-based techniques use information about actual training instances. Although example-based methods have been shown to improve the accuracy of classification tasks over GMMs, this is not the case for recognition tasks. As is well known, speech classification is the task of classifying a speech signal into a given class or type out of a given set of classes or types known a priori, while speech recognition...

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): G10L15/06G10L19/00
CPCG10L15/02
Inventor D·卡涅夫斯基D·纳哈莫B·拉马巴德兰T·N·赛纳斯
Owner INT BUSINESS MASCH CORP
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