Quote:
Originally Posted by LigH
@Warner306: madVR is a renderer. Its purpose is to display video ... conveniently, as much as processing power permits, but most of all: fast enough. Image restoration is not its purpose, only displaying. A video filter for AviSynth or VapourSynth might be the better workflow for neural network based restoration algorithms (similar to NNEDI3). Here it appears quite off-topic to me.
|
Removing artifacts from video is relevant. They all have them. Neural networks are one way to do it. Neural networks can do a lot of things for video. It is up to your imagination, really.