3D Reconstruction of non-visible surfaces of objects from a Single Depth View – Comparative Study
[ 1 ] Instytut Robotyki i Inteligencji Maszynowej, Wydział Automatyki, Robotyki i Elektrotechniki, Politechnika Poznańska | [ D ] phd student | [ P ] employee
[2.2] Automation, electronics, electrical engineering and space technologies
2023
chapter in monograph / paper
english
- robotics
- scene reconstruction
- neural scene representation
EN Scene and object reconstruction is an important problem in robotics, in particular in planning collision-free trajectories or in object manipulation. This paper compares two strategies for the reconstruction of nonvisible parts of the object surface from a single RGB-D camera view. The first method, named DeepSDF predicts the Signed Distance Transform to the object surface for a given point in 3D space. The second method, named MirrorNet reconstructs the occluded objects’ parts by generating images from the other side of the observed object. Experiments performed with objects from the ShapeNet dataset, show that the view-dependent MirrorNet is faster and has smaller reconstruction errors in most categories.
19 - 24
Zaprezentowany na: 4th Polish Conference on Artificial Intelligence PP-RAI'2023, 24-26.04.2023, Łódź, Polska
Licencja Politechniki Łódzkiej
open repository
final published version
at the time of publication
20