zlacker

Is this similar to what GoPro cameras do to remove the selfie-stick? They use video content from adjacent frames to remove the pole and fill in with pixels. I get that the approach here can use imagery that's frames completely differently.