This reminds me of some really cool research in view synthesis! They're able to take a few images from different views and then predict new views. If you're interested, check out the videos and paper here
hal
Also this is a continuation of the previous work by a different group but now they are able to synthesize views without having to meticulously document the viewpoints as one needed to in the previous work.
This reminds me of some really cool research in view synthesis! They're able to take a few images from different views and then predict new views. If you're interested, check out the videos and paper here
Also this is a continuation of the previous work by a different group but now they are able to synthesize views without having to meticulously document the viewpoints as one needed to in the previous work.