Research on Speech Emotion Recognition Technology Based on Machine Learning

In recent years, with the continuous improvement of computer computing power and the emergence of big data processing, artificial intelligence has become an important research field for researchers. The current research on emotion and consciousness is an important part of artificial intelligence research, and the research on speech emotion recognition technology is receiving more and more attention from researchers. Speech emotion recognition has a very broad application prospects. For example, it can be used for student education to identify students’ emotions in time and make appropriate processing,for intelligent human-computer interaction to discover the speaker’s emotional changes in time. It can also be used for criminal investigation to detect the psychological state of the suspect and assist in polygraph detection.This paper mainly studies the MFCC parameters and KNN algorithm of speech signals, and realizes MFCC feature extraction based on MATLAB, and realizes emotion classification through KNN algorithm. The CASIA corpus was used for training and testing, and finally reached an accuracy of 78.00%.