Multi-model fusion Emotion recognition method of speech Expression Based on Deep Learning

As the foundation of human–computer emotional interaction research, emotion recognition affects the development of artificial intelligence technology. At the same time, due to the integration of knowledge of multiple disciplines, the development of emotion recognition research has also led to the rapid development of other disciplines. At present, the research of emotion recognition is mainly concentrated in the field of mono modal emotion recognition such as text, speech, and image. Although unimodal emotion recognition has made many breakthrough achievements, with the passage of time, unimodal emotion recognition has also exposed some problems. It cannot fully describe a certain emotion of the user at the moment, and using multiple modal features to describe a certain emotion together will be more comprehensive and detailed. We humans are well trained because your reading experience recognizes different emotions which make us more reasonable and understandable. But only in the case of a machine, it can easily understand content- based information such as information in text, audio or video, but it is still far behind in accessing the depth of content.