hi, so i am having a really hard time understanding the “depths” and “layers” used in explaining the up convolution part. i think depth 0, depth 1 are used incorrectly
you didn’t mention how you think depth should be used?
so for the down conv part, we have used increasing order, depth 0 (layer 1, layer 2) → depth 1 (layer 1, layer 2). but it all gets a lot messy during the up conv part, where we concating some depth some layer and the applying other operations on other depths, other layers. i have read a lot articles but none seem to help except a possibility that the numbering used there might be incorrect