深層学習ベースの3D Reconstruction / Completionの手法およびデータセットの紹介。過去の有名どころから最新の手法まで。 データセット NYU (ECCV 2012) ShapeNet (arxiv only) ModelNet (CVPR 2015) SUN CG (CVPR 2017) ScanNet (CVPR 2017) Pix3D (CVPR 2018) Scan2CAD (CVPR 2019 Oral) 手法 3D ShapeNets: A Deep Representation for Volumetric Shapes (CVPR 2015) 3D-R2N2 (ECCV 2016) Semantic Scene Completion from a Single Depth Image (CVPR 2017) Pixel2mesh (ECCV 2018
![Deep 3D Reconstruction / Completion](https://cdn-ak-scissors.b.st-hatena.com/image/square/3b7da135f672e3b7184aff89f02a10549421f3f3/height=288;version=1;width=512/https%3A%2F%2Ffiles.speakerdeck.com%2Fpresentations%2F6bf527a179bf4ee4a9f8f754aee8eba4%2Fslide_0.jpg%3F18233102)