To convert a single RGB-D input image into a 3D photo, a team of researchers from Virginia Tech and Facebook developed a deep learning-based image inpainting model that can synthesize color and depth structures in regions occluded in the original view.
Join now