search

cuiziteng / Illumination-Adaptive-Transformer

🌕 [BMVC 2022] You Only Need 90K Parameters to Adapt Light: A Light Weight Transformer for Image Enhancement and Exposure Correction. SOTA for low light enhancement, 0.004 seconds try this for pre-processing.
Apache License 2.0
459 stars 43 forks source link

For grayscale image #19

Closed noringname closed 1 year ago

noringname commented 1 year ago

Hi, thanks for your excellent work, I want to use the model for a grayscale dataset, any suggestions on how to modify the code? It's easy to modify the local branch, but how to modify the global branch, for example, the 3*3 color matrix, how to modify that part?

cuiziteng commented 1 year ago

If the target GT image also 1 channel, maybe you could directly take off colour matrix (because gray scale image don't need colour process information), if the target GT image is 3 channel, this is an interesting image colorization task, and I think you could design several ways for that task to modify the colour matrix (i.e. interpolate colour palette in global branch).