Speech emotion recognition method based on punishment of speaker and independent of speaker

Abstract

The invention discloses a speech emotion recognition method based on punishment of a speaker and independent of the speaker. Speech signal samples are sequentially subjected to pre-processing, original characteristic extraction of speech emotion, dimensionality reduction, and classification judgment of a classifier, wherein in the dimensionality reduction stage, a graph embedding learning method based on the punishment of the speaker is used, and through tag information of the speaker and existing theories on the basis of a graph embedding theory, a combined optimization algorithm is performed for speech signal sample pairs which belong to the same emotion classification but are given by different speakers and speech signal sample pairs which are given by the same speaker but belong to different emotion classifications. Compared with an existing method, in the speech emotion recognition independent of the speaker, the recognition performance of a system can be effectively improved.

Description

technical field [0001] The invention belongs to the field of speech emotion recognition, in particular to a speaker-independent speech emotion recognition method based on speaker punishment. Background technique [0002] With the continuous increase of application requirements, the research of speech emotion recognition (Speech Emotion Recognition, referred to as SER) has been greatly developed in recent years. The results of speech emotion recognition can be applied to automatic analysis and processing of call center corpus, as well as Human-Machine Interaction (HMI) and many other fields to obtain automatic analysis and recognition of speech emotion information and realize machine intelligence. Based on the above requirements, in order to achieve higher system performance, there have been a lot of research work focused on speech emotion recognition. However, a lot of existing work is on how to use expert knowledge or experimental experience to select effective speech emot...

AI Technical Summary

Problems solved by technology

[0006] Technical problem to be solved: Aiming at the deficiencies of the prior art, the present invention proposes a speaker-independent speech emotion recognition method based on speaker penalty, that is, speaker penalty graph learning (Speaker Penalty Graph Learning, referred to as SPGL), which specifically includes The Linear Speaker Penalty Graph Learning Algorithm (LSPGL for short) and the Kernel Speaker Penalty Graph Learning Algorithm (KSPGL for short) solve the problem that in the prior art, speech emotion features are greatly affected by different speakers; and there is a high dimensionality of speech emotion, It is not suitable for feature compression and transmission, which is not conducive to the recognition performance of the system; at the same time, the traditional speaker-independent speech emotion recognition algorithm has poor portability technical problems

Images