MFCC (Mel-frequency cepstral coefficients)

MFCCs are a compact numeric summary of a sound's spectral shape on the mel scale — the dominant audio features of pre-deep-learning speech recognition.

MFCCs distilled each frame to a dozen-odd coefficients capturing vocal-tract character. Modern end-to-end models replaced them with learned features over mel spectrograms, but MFCCs persist across audio analysis and lightweight applications.

Related terms