The RAW images provided to us are 256x256 which is different from what's expected in the final model. So, what should be done here? One immediate thing we could do is resize the RAW images. But I wanted to know from the organizers to get more systematic directions.Posted by: sayak @ Jan. 30, 2021, 5:38 a.m.
Can you please post your questions here: https://ai-benchmark.net/index.php?threads/learned-smartphone-isp-challenge.4/
Sure.Posted by: sayak @ Jan. 31, 2021, 2:52 a.m.
The whole pipeline is like below (assume input resolution is 256x256):
input RAW data (256x256x1) --[deBayer pre-processing]--> deBayer RAW data (128x128x4) --[PUNET model]--> output RGB image (256x256x3)
I guess your confusion is from the deBayer pre-processing.
I will update README and the challenge website for clarification.