Research on Speech Emotion Recognition Based on Teager Energy Operator Coefficients and Inverted MFCC Feature Fusion

As an important part of our daily life, speech has a great impact on the way people communicate. The Mel filter bank used in extraction process MFCC better ability to low-frequency component signal, but it weakens emotional information contained high-frequency signal. We inverted enhance feature processing signal obtain IMFCC coefficients and fuse features order I_MFCC. Finally, more accurately characterize traits, we combined Teager energy operator (TEOC) I_MFCC TEOC&I_MFCC input into CNN_LSTM neural network. Experimental results RAVDESS show that fusion using higher emotion recognition accuracy, system achieves 92.99% weighted accuracy (WA) 92.88% unweighted (UA).