Unlock instant, AI-driven research and patent intelligence for your innovation.

Speech emotional characteristic extraction method for speech emotion recognition

A technology of speech emotion recognition and emotion features, applied in speech recognition, speech analysis, instruments, etc., can solve the problems that cannot meet the requirements of emotion recognition, and achieve the effect of increasing discrimination and improving performance

Active Publication Date: 2012-05-30
INST OF ACOUSTICS CHINESE ACAD OF SCI
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This difference shows that the commonly used feature MFCC based on the uniform Mayer domain cannot meet the requirements of emotion recognition, and there is an urgent need for a feature extraction method that is more suitable for speech emotion recognition.

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 emotional characteristic extraction method for speech emotion recognition
  • Speech emotional characteristic extraction method for speech emotion recognition
  • Speech emotional characteristic extraction method for speech emotion recognition

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0027] The emotion recognition feature calculation method provided by the present invention comprises the following steps:

[0028] 1) Preprocessing and windowing of the speech signal; here, the frame overlap method is adopted. Compared with the traditional MFCC, the frame length and frame shift are 25ms and 10ms respectively. The energy-based voice endpoint detection technology is introduced to judge and discard silence and noise.

[0029] 2) FFT calculation and get uniform passband energy. After the FFT calculation, nbin frequency components are obtained (for 16k sampled speech, approximately 512 sample points and 256 effective values ​​are required), even if the difference between each corresponding point of adjacent frames is very large, the present embodiment adopts uniformly distributed Filter, calculate its passband energy; the number of passband energies is not as small as possible, and its resolution should be guaranteed as much as possible. For the convenience of c...

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 relates to a speech emotional characteristic extraction method for speech emotion recognition, which comprises the following steps of: 1) performing FFT calculation on speech signals to obtain uniform passband energy; 2) by taking the passband energy as a characteristic, calculating an F-Ratio value of speech emotion information at each frequency point to obtain a distribution map of the F-Ratio value of speech emotion information on each frequency band; 3) obtaining a nonuniform Mel subband filter according to the distribution map of the F-Ratio value of speech emotion information on each frequency band, wherein the bandwidth of each passband of the nonuniform Mel subband filter is in inverse proportion to the corresponding F-Ratio value; and 4) acquiring energy output of each subband of a speech signal to be recognized by using the nonuniform Mel subband filter so as to obtain the speech emotional characteristic of the speech signal to be recognized. The speech emotional characteristic extraction method can improve the authentication of multiple classes of emotions, and improves the emotion recognition performance compared with the traditional MFCC characteristic.

Description

technical field [0001] The invention relates to the technical field of speech emotion recognition, in particular, the invention relates to a feature extraction method applied to a speech emotion recognition system. Background technique [0002] With the rapid development of computer network communication technology and multimedia technology, new human-computer interaction technology has become a very active research topic in the field of computer science. The research of speech emotion recognition has important practical significance for enhancing the intelligence and humanization of computers, developing new man-machine environments, and promoting the development of psychology and other disciplines, and will eventually produce great economic and social benefits. Speech emotion recognition is also very beneficial to the improvement of the overall technology of speech recognition. Generally speaking, speech contains rich information such as speech content, speaker, emotion, ...

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): G10L15/02
Inventor 颜永红周瑜孙艳庆李军锋
Owner INST OF ACOUSTICS CHINESE ACAD OF SCI
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More