c2acr#

class diffsptk.CepstrumToAutocorrelation(cep_order: int, acr_order: int, n_fft: int = 512)[source]#

See this page for details.

Parameters:

cep_orderint >= 0: The order of the cepstrum, \(M\).
acr_orderint >= 0: The order of the autocorrelation, \(N\).
n_fftint >> N: The number of FFT bins used for conversion. The accurate conversion requires the large value.

forward(c: Tensor) → Tensor[source]#

Convert cepstrum to autocorrelation.

Parameters:

cTensor [shape=(…, M+1)]: The cepstral coefficients.

Returns:

outTensor [shape=(…, N+1)]: The autocorrelation.

Examples

>>> import diffsptk
>>> import torch
>>> c2acr = diffsptk.CepstrumToAutocorrelation(4, 3)
>>> c = torch.tensor([0.5, -0.3, 0.2, -0.1, 0.05])
>>> r = c2acr(c)
>>> r
tensor([ 3.2404, -1.1920,  0.9037, -0.5838])

diffsptk.functional.c2acr(c: Tensor, acr_order: int, n_fft: int = 512) → Tensor[source]#

Convert cepstrum to autocorrelation.

Parameters:

cTensor [shape=(…, M+1)]: The cepstral coefficients.
acr_orderint >= 0: The order of the autocorrelation, \(N\).
n_fftint >> N: The number of FFT bins used for conversion.

Returns:

outTensor [shape=(…, N+1)]: The autocorrelation.