Speaker-independent speech emotion recognition based on random forest feature selection algorithm

Abstract

Feature selection is a crucial step in the development of a system for identifying emotions in speech. How to select high correlation features is an open question. This paper focuses on feature selection method which aims to extract the most effective acoustic features to improve the performance of emotion recognition. Emotional feature selection of speaker-independent speech based on Random Forest is proposed, which can remove the redundant features that have high correlations with each other. Experiment on the speech emotion recognition based on Support Vector Machines (SVM) is performed, where the speaker-independent features selected by the proposal and Spearman correlation analysis are used for emotion recognition, respectively. Experimental results show that the proposal achieves 78.6% recognition rate on average, which is 2.2% higher than using Spearman.