Leveraging Librosa for Speech Emotion Recognition: Techniques and Applications

Speech emotion recognition (SER), as an important research direction in the field of affective computing, aims to identify and classify emotional states from speech data. In this study, MLP Classifier model was used to classify RAVDESS data set, and the model was trained by extracting audio features such as Mel frequency cepstral coefficient (MFCC), chroma feature and Mell spectrum. The experimental results show that the overall accuracy of the model on this dataset reaches 78.12%. In particular, in the recognition of “calm” emotions, the model performed particularly well, with a recall rate of 82% and an F1 score of 0.87.