levdur#

diffsptk.LevinsonDurbinRecursion#: alias of PseudoLevinsonDurbinRecursion

class diffsptk.PseudoLevinsonDurbinRecursion[source]#

See this page for details. Note that the current implementation does not use the Durbin’s algorithm though the class name includes it.

forward(r)[source]#

Solve a Yule-Walker linear system.

Parameters

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

Returns

aTensor [shape=(…, M+1)]: Gain and LPC coefficients.

Examples

>>> x = torch.randn(5)
tensor([ 0.8226, -0.0284, -0.5715,  0.2127,  0.1217])
>>> acorr = diffsptk.AutocorrelationAnalysis(2, 5)
>>> levdur = diffsptk.LevinsonDurbinRecursion()
>>> a = levdur(acorr(x))
>>> a
tensor([0.8726, 0.1475, 0.5270])