Dimensions for Loss and sdf

[kabbas570]

Hello, @JunMa11 thanks for the nice codes, Can you please clarify what the (x,x,y) is for loss functions and SDF/DT? I am implementing 2D images, so I am confused about depth, height and width.

1- net_output: net logits; shape=(batch_size, class, x, y, z) --> (batch_size, class, height, width, depth) or 2- net_output: net logits; shape=(batch_size, class, x, y, z) --> (batch_size, class, depth,height, width) If I consider the PyTorch required dimensions?

Thanks Cheers Abbas