Voice emotion recognition method in PAD three-dimensional emotion space
A speech emotion recognition and emotion space technology, applied in speech analysis, instruments, etc., can solve problems such as large false recognition rate and the inability of signals to reflect non-stationarity.
Active Publication Date: 2018-10-19
WUHAN UNIV OF TECH
View PDF2 Cites 5 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
MFCC is the Hz spectrum feature calculated by using the nonlinear correspondence between Mel frequency and Hz frequency. The Mel frequency scale can strengthen the low-frequency details of the speech signal, so MFCC can highlight the useful information of the speech signal and reduce the impact of environmental noise on the speech signal. Interference can effectively identify speech emotion, but because the non-stationarity of emotional speech signals is particularly obvious, the direct FFT of the signal cannot reflect its non-stationarity, so using MFCC alone for speech emotion recognition has a large misrecognition rate
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0014] In order to facilitate those of ordinary skill in the art to understand and implement the present invention, the present invention will be described in further detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the implementation examples described here are only used to illustrate and explain the present invention, and are not intended to limit this invention.
[0015] The CASIA Chinese emotional corpus used in this embodiment is a discrete speech emotional database developed by the Chinese Academy of Sciences Automation, but due to a completed research project, the Institute of Psychology, Chinese Academy of Sciences hired 346 college students to use its revised simplified version of the Sinicization The PAD emotion scale evaluates the PAD value of 14 specific emotional categories, and obtains the values of these 14 emotions in P (pleasure), A (activation) and D (dominance), which includes CASIA There are 6 types ...
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
Login to View More Abstract
The invention discloses a voice emotion recognition method in a PAD three-dimensional emotion space, a PAD three-dimensional emotion model based on a dimensionality theory is selected and used as a recognition result manifestation mode, Mel-frequency cepstral coefficient, time ignition sequence and ignition position information characteristics are used for PAD value prediction of voice emotion separately, a correlation analysis is made from three dimensions of P (pleasure), A (activation) and D (dominance), weight coefficients of the three characteristics are calculated, and a final predictedvalue of voice emotion in PAD three-dimensional emotion space is obtained by weighted fusion. Experiments show that the method can locate the emotional state of voice in the emotion space more meticulously, pay more attention to the expression and reflection of the internal components of emotion, and reflect the polarity and degree of emotional expression more appropriately, thereby displaying mixed emotion content in emotion voice.
Description
technical field [0001] The invention belongs to the field of speech emotion recognition, and relates to a speech emotion recognition method, in particular to a speech emotion recognition method in a PAD three-dimensional emotion space. Background technique [0002] In the field of speech emotion recognition, commonly used cepstral features generally include Mel-Frequency Cepstral Coefficient (MFCC) and the like. MFCC is the Hz spectrum feature calculated by using the nonlinear correspondence between Mel frequency and Hz frequency. The Mel frequency scale can strengthen the low-frequency details of the speech signal, so MFCC can highlight the useful information of the speech signal and reduce the impact of environmental noise on the speech signal. Interference can effectively identify speech emotion, but because the non-stationarity of emotional speech signals is particularly obvious, the direct FFT of the signal cannot reflect its non-stationarity, so using MFCC alone for sp...
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
Login to View More IPC IPC(8): G10L25/03G10L25/24G10L25/27G10L25/63
CPCG10L25/03G10L25/24G10L25/27G10L25/63
Inventor 程艳芬陈逸灵李超
Owner WUHAN UNIV OF TECH