ShapeClipper: Scalable 3D Shape Learning from Single-View Images via Geometric and CLIP-Based Consistency

Zixuan Huang; Varun Jampani; Anh Thai; Yuanzhen Li; Stefan Stojanov; James M. Rehg

doi:10.1109/cvpr52729.2023.01241

CVPR2023顶会

ShapeClipper: Scalable 3D Shape Learning from Single-View Images via Geometric and CLIP-Based Consistency

Zixuan Huang, Varun Jampani, Anh Thai, Yuanzhen Li, Stefan Stojanov, James M. Rehg

arXiv DOI 出版方

2023年份

4顶会引用

摘要

Google Research Figure 1. We propose a method that reconstructs 3D object shape from single view real-world images. Our method learns high-quality reconstruction through single-view supervision without known viewpoint and can reconstruct shapes of various objects.