Method for recognizing speaker based on multivariate core logistic regression model

A regression model and implementation method technology, applied in speech analysis, instruments, etc., can solve problems such as slow speed, low recognition rate, and complex model construction

Inactive Publication Date: 2011-11-23
ZHEJIANG UNIV OF TECH
View PDF7 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In order to overcome the shortcomings of low recognition rate, complex model construction and slow speed of the existing speaker identification implementation methods, the present invention provides a multivariate kernel logistic regression model based on high recognition rate, simple model construction, and good rapidity The implementation method of speaker identification

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
  • Method for recognizing speaker based on multivariate core logistic regression model
  • Method for recognizing speaker based on multivariate core logistic regression model
  • Method for recognizing speaker based on multivariate core logistic regression model

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0042] The present invention will be further described below.

[0043] A method for implementing speaker discrimination based on a multivariate kernel logistic regression model, comprising the following steps:

[0044] A), speaker speech feature extraction: collect the speech signal of the speaker to be identified, and carry out preprocessing; then extract the Mel cepstrum parameters, the Mel cepstrum parameters are 13th order cepstrum parameters, which will describe the speaker's personality characteristics The weaker zeroth order coefficient is removed, and the remaining 12-dimensional feature vector is used as the speaker identification input vector;

[0045]B), speaker model construction: multivariate kernel logistic regression model is used as the speaker identification model,

[0046] p ( c i = k | x ‾ ; ...

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 method for recognizing a speaker based on a multivariate core logistic regression model, comprising the following steps: (A) extracting voice features of the speaker: collecting voice signals of the speaker to be recognized to pre-process, and then extracting mel cepstrum parameters; (B) constructing a speaker model: using a multivariate core logistic regression model asa speaker recognition model; (C) training the speaker recognition model: using the feature vectors extracted from the step A as input training samples, through a minimal sequence optimization algorithm, carrying out an iterative training to optimize the model parameters; (D) recognizing the speaker: extracting the feature vectors of the voice signals of the speaker to be recognized and inputting to the recognition model of the trained speaker, and giving out a posterior probability of each speaker by the multivariate core logistic regression model, wherein the highest probability value is a recognition result. The invention has high rate of recognition, simple model construction and good rapidity.

Description

technical field [0001] The invention relates to the fields of signal processing, machine learning and pattern recognition, in particular to a method for realizing speaker identification. Background technique [0002] Speaker identification refers to automatically identifying whether the speaker is in the specified speaker set by analyzing and processing the speaker's voice signal in a limited set and extracting features, and then confirming the specific identity of the speaker. The basic principle of speaker identification is to build a classification model for each speaker that can describe its personality characteristics. Therefore, excellent model construction is one of the key technologies for speaker identification. [0003] Traditional speaker identification models include generative models such as Mixed Gaussian Model (GMM) and Hidden Markov Model (HMM). Although these models can achieve good recognition efficiency, a large number of training samples are required to...

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): G10L17/00G10L17/04
Inventor 王万良郑建炜郑泽萍韩姗姗蒋一波王震宇王磊陈胜勇
Owner ZHEJIANG UNIV OF TECH
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