1 paper
Olli Niemitalo, Otto Rosenberg, Nathaniel Narra +2
We present linear prediction as a differentiable padding method. For each channel, a stochastic autoregressive linear model is fitted to the padding input by minimizing its noise t…