GFCC¶

standard mode | Spectral category

Inputs¶

spectrum (vector_real) - the audio spectrum

Outputs¶

bands (vector_real) - the energies in ERB bands

gfcc (vector_real) - the gammatone feature cepstrum coefficients

Parameters¶

dctType (integer ∈ [2, 3], default = 2) :
the DCT type

highFrequencyBound (real ∈ (0, ∞), default = 22050) :
the upper bound of the frequency range [Hz]

inputSize (integer ∈ (1, ∞), default = 1025) :
the size of input spectrum

logType (string ∈ {natural, dbpow, dbamp, log}, default = dbamp) :
logarithmic compression type. Use ‘dbpow’ if working with power and ‘dbamp’ if working with magnitudes

lowFrequencyBound (real ∈ [0, ∞), default = 40) :
the lower bound of the frequency range [Hz]

numberBands (integer ∈ [1, ∞), default = 40) :
the number of bands in the filter

numberCoefficients (integer ∈ [1, ∞), default = 13) :
the number of output cepstrum coefficients

sampleRate (real ∈ (0, ∞), default = 44100) :
the sampling rate of the audio signal [Hz]

silenceThreshold (real ∈ (0, ∞), default = 1e-10) :
silence threshold for computing log-energy bands

type (string ∈ {magnitude, power}, default = power) :
use magnitude or power spectrum

Description¶

This algorithm computes the Gammatone-frequency cepstral coefficients of a spectrum. This is an equivalent of MFCCs, but using a gammatone filterbank (ERBBands) scaled on an Equivalent Rectangular Bandwidth (ERB) scale.

References:: [1] Y. Shao, Z. Jin, D. Wang, and S. Srinivasan, “An auditory-based feature for robust speech recognition,” in IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP’09), 2009, pp. 4625-4628.

Source code¶

C++ source code

C++ header file