CLI/automation for scaling wallpapers to monitor resolutions

Read monitor info to generate config file ->
User inputs relevant folders with wallpapers + Set up paths for de-JPG artifact removal (optional) and upscaling models ->
Run main.py ->
Program checks for files already at target resolution then ignores them ->
Program checks for files at resolutions larger than target resolution and stores them in a separate bucket ->
Program checks for files already at correct aspect ratio and puts them in a separate bucket ->
Images not at target aspect ratio get run through a Computer Vision model that looks for the most salient region and crops to target aspect ratio, using the salient region as guideline (salient region on left means crop more on the right) ->
Images get run through AI Preprocessing+Upscaling model on GPU (with CPU fallback) ->
All image buckets get merged ->
Images are downscaled to target resolution using ->
Images get saved to a new folder in parent folder with a suffix

For people who likes their wallpapers to rotate and want to be able to use pictures of any quality as their wallpaper