• Jul 2, 2026 Extracting Mfcc Features For Emotion Recognition From subtraction or Wiener filtering are essential particularly with noisy audio recordings Normalization Normalizing the MFCCs eg using zscore normalization can significantly improve the performance of machine learning models by scaling th BY Beatrice Johnson