Reconstruction by Generation: 3D Multi-Object Scene Reconstruction from Sparse Observations

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zadaianchuk, Andrii, Barcellona, Leonardo, Schuenemann, Lennard, Gumbsch, Christian, Wang, Zehao, Irshad, Muhammad Zubair, Despinoy, Fabien, Aljundi, Rahaf, Gavves, Stratis, Zakharov, Sergey
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909001379217408
author Zadaianchuk, Andrii
Barcellona, Leonardo
Schuenemann, Lennard
Gumbsch, Christian
Wang, Zehao
Irshad, Muhammad Zubair
Despinoy, Fabien
Aljundi, Rahaf
Gavves, Stratis
Zakharov, Sergey
author_facet Zadaianchuk, Andrii
Barcellona, Leonardo
Schuenemann, Lennard
Gumbsch, Christian
Wang, Zehao
Irshad, Muhammad Zubair
Despinoy, Fabien
Aljundi, Rahaf
Gavves, Stratis
Zakharov, Sergey
contents Accurately reconstructing complex full multi-object scenes from sparse observations remains a core challenge in computer vision and a key step toward scalable and reliable simulation for robotics. In this work, we introduce RecGen, a generative framework for probabilistic joint estimation of object and part shapes, as well as their pose under occlusion and partial visibility from one or multiple RGB-D images. By leveraging compositional synthetic scene generation and strong 3D shape priors, RecGen generalizes across diverse object types and real-world environments. RecGen achieves state-of-the-art performance on complex, heavily occluded datasets, robustly handling severe occlusions, symmetric objects, object parts, and intricate geometry and texture. Despite using nearly 80% fewer training meshes than the previous state of the art SAM3D, RecGen outperforms it by 30.1% in geometric shape quality, 9.1% in texture reconstruction, and 33.9% in pose estimation.
format Preprint
id arxiv_https___arxiv_org_abs_2604_27106
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Reconstruction by Generation: 3D Multi-Object Scene Reconstruction from Sparse Observations
Zadaianchuk, Andrii
Barcellona, Leonardo
Schuenemann, Lennard
Gumbsch, Christian
Wang, Zehao
Irshad, Muhammad Zubair
Despinoy, Fabien
Aljundi, Rahaf
Gavves, Stratis
Zakharov, Sergey
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Robotics
Accurately reconstructing complex full multi-object scenes from sparse observations remains a core challenge in computer vision and a key step toward scalable and reliable simulation for robotics. In this work, we introduce RecGen, a generative framework for probabilistic joint estimation of object and part shapes, as well as their pose under occlusion and partial visibility from one or multiple RGB-D images. By leveraging compositional synthetic scene generation and strong 3D shape priors, RecGen generalizes across diverse object types and real-world environments. RecGen achieves state-of-the-art performance on complex, heavily occluded datasets, robustly handling severe occlusions, symmetric objects, object parts, and intricate geometry and texture. Despite using nearly 80% fewer training meshes than the previous state of the art SAM3D, RecGen outperforms it by 30.1% in geometric shape quality, 9.1% in texture reconstruction, and 33.9% in pose estimation.
title Reconstruction by Generation: 3D Multi-Object Scene Reconstruction from Sparse Observations
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Robotics
url https://arxiv.org/abs/2604.27106