Extraction and modeling method for Chinese speech sensibility information

Abstract

The invention provides a method for extracting and modeling the emotional information of a Chinese sound; the extracting method for the emotional information of the Chinese sound is that: formulate the specification of a emotional speech database, which includes the pronouncer specification, the recording play book design specification and the naming specification of audio files and so on; collect the emotional speech data; evaluate the validity of the emotional speech, namely, at least ten evaluators apart from a speaker carry out a subjective listen evaluation experiment on the emotional speech data. The modeling method of the emotional information of the Chinese sound is that: extract the emotional characteristics of the sound, define and distinguish the characteristic combination of each emotion type; adopt different characteristic combinations to train the SVM model of a multilevel sound emotion recognition system; verify the identification effect of the classifying models, namely, verify the classification effect of the multilevel classification models of sound emotion in a situation unrelated to the speaker by adopting a cross leave-one-out method. The method solves the problems that the domestic emotional speech databases are less in emotion type and the number of the domestic emotional speech database is very limited; at the same time, the method realizes an efficient speech emotion identification system.

Description

(1) Technical field: [0001] The invention relates to a method for extracting and modeling Chinese speech emotion information, and belongs to the field of information technology. (2) Background technology: [0002] The research of automatic speech emotion recognition technology mainly involves two aspects: one is to establish a high-quality emotional speech database to provide necessary data for speech emotion research; the other is to establish an efficient speech emotion recognition model to quickly and effectively recognize the emotion of speech signals status. [0003] To establish an emotional speech database, firstly, the categories and classification methods of emotions must be established. In most research methods, researchers use everyday language tags to identify and classify emotions, that is, describe emotion classification as a discrete model. Emotional speech data can be divided into three categories: natural, performance, and elicitation according to the naturalness...

AI Technical Summary

Problems solved by technology

Before the present invention, the emotional features mainly used by researchers included prosodic features and their derived parameters, such as time, energy, gene frequency, formant, etc. However, the relationship between these parameters and emotional expression is very complicated, and different emotional divisions often It is suitable to adopt a specific combination of features, and there is little research on it in the past literature

The speech signal is a typical non-stationary signal, and the linear models adopted by the K-nearest neighbor method and the principal component analysis method are too simple to obtain a good recognition rate; the establishment and training of the hidden Markov model (HMM) takes too long, In practice, it is necessary to solve the problem of high computational complexity; the artificial neural network has a high degree of nonlinear modeling and strong classification ability, but it cannot overcome the local minimum problem

In addition, the appeal recognition methods all belong to the category of statistical machine learning. According to theoretical analysis, the recognition performance can only be guaranteed when the number of training samples tends to infinity, but there is always the problem of small sample learning in practical applications.

Images