Closed learningyan closed 3 years ago
To clarify, I think you are asking for DAGF models which transform clean images to Glass, TOLED, POLED (which we refer to as "simulation" in our paper).
Here are the checkpoints:
Model | Config | Checkpoint |
---|---|---|
Clean to Glass | guided_filter_l1_tanh_pixelshuffle_forward_glass | clean2glass |
Clean to POLED | guided_filter_l1_tanh_pixelshuffle_forward_poled | clean2poled |
Clean to TOLED | guided_filter_l1_tanh_pixelshuffle_forward_toled | clean2toled |
The code for these models can be found in the experiments branch.
We'll add this to the README too.
awesome work !
According to your paper , you trained a simulation network to generate simulated data. I want to reproduce this work on other datasets, but I cannot find the simulate pretrained checkpoints. Cau you provide pretrained checkpoints of simulating poled or toled data ? thanks!