Audio features for analysis
Local feature extraction: RMS, zero-crossing rate, spectral centroid, 85% energy rolloff, spectral flatness, spectral flux and thirteen mel-frequency cepstral coefficients from an own DFT, mel filter bank and DCT-II.
Local feature extraction: RMS, zero-crossing rate, spectral centroid, 85% energy rolloff, spectral flatness, spectral flux and thirteen mel-frequency cepstral coefficients from an own DFT, mel filter bank and DCT-II.