Since I assumed that all VSGAN based stuff is spatial only and only works basically on single images, but since a few folks mentioned that the models would take into account temporal information when the model is applied I wanted to ask whether this is true. :)
So does VSGAN use temporal data when using different models?
It would depend on the architecture used. EGVSR is the only architecture currently supported that does anything like that but is a hell of a lot harder and longer to train.
Since I assumed that all VSGAN based stuff is spatial only and only works basically on single images, but since a few folks mentioned that the models would take into account temporal information when the model is applied I wanted to ask whether this is true. :) So does VSGAN use temporal data when using different models?