DISN: Deep Implicit Surface Network for High-quality Single-view 3D Reconstruction - Jianfei Guo

目录

目录

DISN: Deep Implicit Surface Network for High-quality Single-view 3D Reconstruction

Jianfei Guo 出版于 paper_reading

2020-12-09 2020-12-09 约 123 字预计阅读 1 分钟

<DISN> DISN: Deep implicit surface network for high-quality single-view 3D reconstruction

NeruIPS2019 Advances in Neural Information Processing Systems

Qiangeng Xu, Weiyue Wang, Duygu Ceylan, Radomir Mech, Ulrich Neumann

University of Southern California Adobe

SDF, single-view, encoder

目录

编者按

训练时是有3D shape 的SDF的真值数据的；图像feature只是提供一个辅助的code输入而已

Motivation

希望学到的shape，不仅全局特征好，还想有局部fine grained details 细粒度细节

overview

同时用global features和local features来infer SDF