Complete Object-Compositional Neural Implicit Surfaces With 3D Pseudo Supervision

Neural implicit surface reconstruction has recently emerged as a prominent paradigm in multi-view 3D reconstruction using deep learning. In contrast to traditional multi-view stereo methods, signed distance function (SDF)-based approaches leverage neural networks to effectively represent 3D scenes....

Full description

Saved in:
Bibliographic Details
Main Authors: Wongyeom Kim, Jisun Park, Kyungeun Cho
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10900377/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850023459432693760
author Wongyeom Kim
Jisun Park
Kyungeun Cho
author_facet Wongyeom Kim
Jisun Park
Kyungeun Cho
author_sort Wongyeom Kim
collection DOAJ
description Neural implicit surface reconstruction has recently emerged as a prominent paradigm in multi-view 3D reconstruction using deep learning. In contrast to traditional multi-view stereo methods, signed distance function (SDF)-based approaches leverage neural networks to effectively represent 3D scenes. Furthermore, to reconstruct scenes and individual objects separately, some studies have extended the framework for object-compositional neural implicit surface reconstruction by utilizing 2D instance masks to supervise the SDF of each object. Nonetheless, these methods often reconstruct objects as partial shapes in scenes captured from sparse viewpoints or in complex scenes containing multiple objects. This issue primarily stems from the absence of a 3D prior, which fails to provide sufficient geometry for partially observed and occluded regions. We propose a framework for completing the partial object shapes of object-compositional neural implicit representation utilizing a diffusion-based 3D mesh generation model. The existing diffusion model, trained only on large-scale 3D object datasets, generates complete shapes from partial shapes; however, their results differ significantly from the objects in the scene. To complete the representation of partial shapes while ensuring shape consistency across multi-view images, we combine the SDF values, output by the diffusion model, with the object-compositional neural implicit representation. The combined representation is then volume-rendered to fine-tune the diffusion model utilizing a 2D prior. Furthermore, the complete shape generated by our method can serve as pseudo 3D priors to provide the geometry for the unobserved regions in object-compositional representation. Extensive experiments demonstrate that our novel framework significantly improves the reconstruction quality of unobserved regions.
format Article
id doaj-art-1ed0f8a777d147b78f4610afa1a90eb4
institution DOAJ
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-1ed0f8a777d147b78f4610afa1a90eb42025-08-20T03:01:22ZengIEEEIEEE Access2169-35362025-01-0113361513616110.1109/ACCESS.2025.354470510900377Complete Object-Compositional Neural Implicit Surfaces With 3D Pseudo SupervisionWongyeom Kim0https://orcid.org/0009-0008-4788-935XJisun Park1https://orcid.org/0000-0002-4304-1780Kyungeun Cho2https://orcid.org/0000-0003-2219-0848Department of Computer Science and Artificial Intelligence, Dongguk University, Jung-gu, Seoul, Republic of KoreaNUI/NUX Platform Research Center, Dongguk University, Jung-gu, Seoul, Republic of KoreaDivision of AI Software Convergence, Dongguk University, Jung-gu, Seoul, Republic of KoreaNeural implicit surface reconstruction has recently emerged as a prominent paradigm in multi-view 3D reconstruction using deep learning. In contrast to traditional multi-view stereo methods, signed distance function (SDF)-based approaches leverage neural networks to effectively represent 3D scenes. Furthermore, to reconstruct scenes and individual objects separately, some studies have extended the framework for object-compositional neural implicit surface reconstruction by utilizing 2D instance masks to supervise the SDF of each object. Nonetheless, these methods often reconstruct objects as partial shapes in scenes captured from sparse viewpoints or in complex scenes containing multiple objects. This issue primarily stems from the absence of a 3D prior, which fails to provide sufficient geometry for partially observed and occluded regions. We propose a framework for completing the partial object shapes of object-compositional neural implicit representation utilizing a diffusion-based 3D mesh generation model. The existing diffusion model, trained only on large-scale 3D object datasets, generates complete shapes from partial shapes; however, their results differ significantly from the objects in the scene. To complete the representation of partial shapes while ensuring shape consistency across multi-view images, we combine the SDF values, output by the diffusion model, with the object-compositional neural implicit representation. The combined representation is then volume-rendered to fine-tune the diffusion model utilizing a 2D prior. Furthermore, the complete shape generated by our method can serve as pseudo 3D priors to provide the geometry for the unobserved regions in object-compositional representation. Extensive experiments demonstrate that our novel framework significantly improves the reconstruction quality of unobserved regions.https://ieeexplore.ieee.org/document/10900377/Deep learningmesh generationsurface reconstruction
spellingShingle Wongyeom Kim
Jisun Park
Kyungeun Cho
Complete Object-Compositional Neural Implicit Surfaces With 3D Pseudo Supervision
IEEE Access
Deep learning
mesh generation
surface reconstruction
title Complete Object-Compositional Neural Implicit Surfaces With 3D Pseudo Supervision
title_full Complete Object-Compositional Neural Implicit Surfaces With 3D Pseudo Supervision
title_fullStr Complete Object-Compositional Neural Implicit Surfaces With 3D Pseudo Supervision
title_full_unstemmed Complete Object-Compositional Neural Implicit Surfaces With 3D Pseudo Supervision
title_short Complete Object-Compositional Neural Implicit Surfaces With 3D Pseudo Supervision
title_sort complete object compositional neural implicit surfaces with 3d pseudo supervision
topic Deep learning
mesh generation
surface reconstruction
url https://ieeexplore.ieee.org/document/10900377/
work_keys_str_mv AT wongyeomkim completeobjectcompositionalneuralimplicitsurfaceswith3dpseudosupervision
AT jisunpark completeobjectcompositionalneuralimplicitsurfaceswith3dpseudosupervision
AT kyungeuncho completeobjectcompositionalneuralimplicitsurfaceswith3dpseudosupervision