search

sh4174 / 3DStyleGAN

3D StyleGAN2 for Medical Images
63 stars 16 forks source link

could not broadcast input array from shape (24,1,256,256,256) into shape (24,1,255,306,357) #5

Open lx123-j opened 1 year ago

lx123-j commented 1 year ago

could you tell me what should I do if my image are 256×256×256

sebasmos commented 1 year ago

They downsample the image in this Line. This was configured for a shape (160×192×224). You may need to change it to you input shape

ans92 commented 1 year ago

Hi @sebasmos and @lx123-j , I want to know how we can change the shape of second dimension i.e in above case (24,1,255,306,357). We can change the the first shape through @sebasmos answer but still the second shape remains there. I also want to change that. So could any one of you help me in this regard. I would be very thankful to you.

ans92 commented 1 year ago

The answer of above is: https://github.com/sh4174/3DStyleGAN/blob/dce413edecc5ab01b99cb895885bb02ae9a37875/training/dataset.py#LL146C13-L146C13

newsblue commented 1 month ago

could you tell me what should I do if my image are 256×256×256

need to generate tfrecord file first, but length width depth must in the order of 2

newsblue commented 1 month ago

could you tell me what should I do if my image are 256×256×256

I am researching