0538

In the paper, you have mentioned that each convolutional block for the DHAN follows by an Instance Normalization (IN) and LeakyReLU layer, however, it uses Batch Normalization here instead. https://github.com/vinthony/ghost-free-shadow-removal/blob/93cc1d68c1870def781f4aabc208a38d0869766e/networks.py#L25...

Questions of specific constants

1

What do the 2.6, 4.8, 3.7, etc. mean? https://github.com/ceciliavision/perceptual-reflection-removal/blob/1d2449463ab9e436443204f3a00eb75740e9a8f9/main.py#L189

Location Reset

1

I was able to successfully set a location at first, but the location goes away / reset after some time (unplugged from computer). Device: iPhone 12 Pro Max iOS: 15.2.1...

Dumb Question about DM vs LDM

2

Do they only differ between the use of VAE to encode the inputs into embedding (and the conditional input part)? So if I wanted to make this in latent space,...

On architecture drawing

1

Just a random non-technical question, which software did you use for drawing the model architecture? It looks super sleek.

What are the recommended settings (number of epochs, etc.) if we were to train from scratch?