3D Photography Using Context-Aware Layered Depth Inpainting

Meng-Li Shih, Shih-Yang Su, Johannes Kopf, Jia-Bin Huang
2020 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
https://shihmengli.github.io/3D-Photo-Inpainting (a) Depth-warping (holes) (b) Depth-warping (stretching) (c) Facebook 3D photo (d) Our result Figure 1. 3D photography from a single RGB-D image. Naïve methods either produce holes (a) or stretch content (b) at disocclusions. Color and depth inpainting using diffusion is better, but provides a too smooth appearance (c). Our approach is capable of synthesizing new color/depth texture and structures, leading to more photorealistic novel views (d).
doi:10.1109/cvpr42600.2020.00805 dblp:conf/cvpr/ShihSKH20 fatcat:u75konk4xnasfcyprhpxwigtva