Re:NeRF in the Wild: reconstructing 3D scenes from internet ...
by airstrike on Thursday August 06, 2020 @05:16PM
(#24075550)
But if we know what a car ought to look like in 3D, can't we take the one photo we have from one direction and just fill in the blanks with that a priori 3D knowledge?
[ Parent
| on HN ]
Re:NeRF in the Wild: reconstructing 3D scenes from internet ...
by withjive on Thursday August 06, 2020 @06:41PM
(#24076190)
Similar to how GPT-3 can be applied not only to create Text, but also fill in missing pieces of Images (ie. complete the missing half of a face).
Would the logical next step, use GPT-3 to create a 3D world? :)
Re:NeRF in the Wild: reconstructing 3D scenes from internet ...
by airstrike on Thursday August 06, 2020 @11:45PM
(#24077964)
GPT-3D rolls off the tongue nicely
Re:NeRF in the Wild: reconstructing 3D scenes from internet ...
by jhurliman on Thursday August 06, 2020 @11:38PM
(#24077895)
Getting there. This is one part of the puzzle: https://arxiv.org/pdf/2007.11965