c2acr#

class diffsptk.CepstrumToAutocorrelation(acr_order, fft_length)[source]#

See this page for details.

Parameters:

acr_orderint >= 0 [scalar]: Order of autocorrelation, \(M_2\).
fft_lengthint >= 2 [scalar]: Number of FFT bins, \(L\).

forward(c)[source]#

Convert cepstrum to autocorrelation.

Parameters:

cTensor [shape=(…, M1+1)]: Cepstrum.

Returns:

rTensor [shape=(…, M2+1)]: Autocorrelation.

Examples

>>> c = diffsptk.nrand(4)
>>> c
tensor([-0.1751,  0.1950, -0.3211,  0.3523, -0.5453])
>>> c2acr = diffsptk.CepstrumToAutocorrelation(4, 16)
>>> r = c2acr(c)
>>> r
tensor([ 1.0672, -0.0485, -0.1564,  0.2666, -0.4551])