I’m trying to implement some customized Convolution and I need the im2col function
I checked the implementation in SpatialDilatedConvolution.c
and the im2col is written as:
THNN_(col2im)(
THTensor_(data)(gradColumns),
nInputPlane, inputHeight, inputWidth, outputHeight, outputWidth,
kH, kW, padH, padW, dH, dW,
dilationH, dilationW,
THTensor_(data)(gradInput_n)
);
Since I’m just using Float type, so I tried THNN_Floatim2col
, but it does not work
My questions are:
-
How can I call
im2col
in c extensions with Float type (in GPU and CPU version)? -
Do I need to include any head files?