CVPR2020

FroDO: From Detections to 3D Objects

Martin Rünz, Kejie Li, Meng Tang, Lingni Ma, Chen Kong, Tanner Schmidt, Ian D. Reid, Lourdes Agapito, Julian Straub, Steven Lovegrove, Richard A. Newcombe

DOI 出版方

摘要

Figure 1: Given a localized input RGB sequence, FroDO dectects objects and infers their pose and a progressively fine grained and expressive object shape representation. Results on a real-world sequence from ScanNet [7].