uzh-rpg/rpg_e2vid

Seeding inference with an image

Opened this issue · 0 comments

bradfeiger commented

Is there a way to seed the video reconstruction by inputting a framed image at the first time step during inference?