Hello,

I've seen people on the OpenModelDB website sharing PyTorch files (mostly) that allow them to filter, or even restore, videos using artificial intelligence.

I'd like to know if it's difficult to create a model, as I've tested several and noticed that the results depend heavily on the videos being filtered.

Please note that I'm a complete beginner, and this will be my first time undertaking such a project.

Until now, I've only filtered videos using AviSynth and Megui.

In that regard, if upscaling isn't an option, is creating an AI model still worthwhile?

Thank you in advance for your help.