Improving textures

Video AI is quite good at improving edges, removing noise and compression artifacts, but it often lacks the ability to recover all sorts of surfaces, textures, faces, and text especially on very low quality video footage. I would love to see improvements here. The new de-blur option is a major step forward in this regard, but even then surfaces are often not recovored very well. The AI seems to have no idea what to do with those. :eyes: