If you look in how the 1d and 2d function work, could can probably extend it to work for 3d. This is based on convolution and we have 3d convolution on Theano. Do it should work I think.
Fred Le 28 sept. 2016 09:18, "Ioannis Marras" <[email protected]> a écrit : > To explain it better, I have a 3D tensor and I have to upsample it using > different ratio per axis. How can I do this? I am aware of the function '' > bilinear_upsampling'' wich works for 1 or 2 dimensions. In my case, I > have 3 axes so if I use 3 times the same function for each axis it is very > slow... > > On Wednesday, 28 September 2016 13:57:02 UTC+1, Ioannis Marras wrote: >> >> Hi all, >> >> I would like to use the MATLAB function ''interpn(Data,Xloc,Yloc,Zloc,Rloc)'' >> in Theano. >> Do you know how can I do this efficiently? >> Has someone done it before so he/she can help me on this? >> >> Any suggestions are welcome. >> >> Many Thanks, >> >> Ioannis >> > -- > > --- > You received this message because you are subscribed to the Google Groups > "theano-users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/d/optout. > -- --- You received this message because you are subscribed to the Google Groups "theano-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
