Voice motion recognition method based on feature selection and optimization

Abstract

The invention discloses a voice emotion recognition method based on feature selection and optimization, and mainly solves the problem of low accuracy rate of speech emotion recognition in the prior art. According to the scheme, the method includes the following steps of: 1) establishing an emotion voice database and obtaining original voice; 2) carrying out preemphasis, windowing by frames and end-point detection and other preprocessing on the original voice; 3) extracting emotion characteristic parameters of the pre-treated voice; 4) selecting the optimal parameters of the voice emotion by using a random forest algorithm; 5) inputting the optimal voice emotion characteristic parameters into a trained convolution neural network to obtain the result of the voice emotion recognition. By analyzing the importance of the voice emotion characteristics, the optimal voice emotion characteristic parameters are obtained, the accuracy of the convolution neural network algorithm to voice emotion recognition is improved, and the method can be applied to mobile phone communication, human-computer interaction and recognition of speaker emotion in medical diagnosis and criminal investigation.

Description

Technical field [0001] The invention belongs to the technical field of voice signal processing, and particularly relates to a voice emotion recognition method, which can be used for mobile phone communication, human-computer interaction, medical diagnosis and criminal investigation. Background technique [0002] The earliest research related to speech emotion recognition appeared in the mid-1980s. This is the first time that humans use acoustic features to study and analyze emotional features. In 1999, Moriyama proposed that there is a linear correlation model between speech and emotion, and realized the initial application of speech emotion in e-commerce. Since the beginning of the 21st century, the research of speech emotion recognition has been developed in various fields and gradually applied to the fields of human-computer interaction, mobile phone communication and clinical medicine, which has attracted worldwide attention. [0003] Speech emotion recognition is the process ...

AI Technical Summary

Problems solved by technology

For example, in the emotion recognition method of Mandarin speech based on PCA and SVM, its emotion feature selection algorithm is principal component analysis PCA, and multiple SVMs are used for emotion recognition of speech, because the PCA algorithm is not suitable for feature selection of samples with non-Gaussian distribution , so this support vector machine SVM speech emotion recognition method is only suitable for the case of small sample size, for the case of large sample size, the kernel function mapping dimension is very high, the calculation complexity is high, and the speech emotion recognition rate is low, and for The choice of kernel function for nonlinear problems is difficult because there is no general standard

As another example, the convolutional neural network CNN speech emotion recognition method that has appeared in recent years, because it only uses spectral features such as Mel-frequency cepstral coefficients MFCC and spectrograms, without adding voice quality features and prosody features, it still cannot be obtained. More accurate speech emotion recognition rate

Images