A Multimodal Emotion Recognition Method Based on Speech-Text

Abstract

Aiming at the problems of low recognition rate and easy to be disturbed by noise in the process of single-mode speech emotion recognition, this paper proposes a speech emotion analysis method based on multi feature fusion of speech and semantics. This method uses opensmile to extract acoustic features and Bi long and short term memory network (Bi-LSTM) to extract semantic features, then carries out feature data fusion, and then inputs the fused data into SVM classification model to obtain the final emotion classification result. This method can effectively solve the shortcomings of single-mode emotion recognition and improve the efficiency and accuracy of recognition.