DeOcc-1-to-3: 3D De-Occlusion from a Single Image via Self-Supervised Multi-View Diffusion

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Qu, Yansong, Dai, Shaohui, Li, Xinyang, Wang, Yuze, Shen, You, Cao, Liujuan, Ji, Rongrong
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912457183723520
author Qu, Yansong
Dai, Shaohui
Li, Xinyang
Wang, Yuze
Shen, You
Cao, Liujuan
Ji, Rongrong
author_facet Qu, Yansong
Dai, Shaohui
Li, Xinyang
Wang, Yuze
Shen, You
Cao, Liujuan
Ji, Rongrong
contents Reconstructing 3D objects from a single image remains challenging, especially under real-world occlusions. While recent diffusion-based view synthesis models can generate consistent novel views from a single RGB image, they typically assume fully visible inputs and fail when parts of the object are occluded, resulting in degraded 3D reconstruction quality. We propose DeOcc-1-to-3, an end-to-end framework for occlusion-aware multi-view generation that synthesizes six structurally consistent novel views directly from a single occluded image, enabling reliable 3D reconstruction without prior inpainting or manual annotations. Our self-supervised training pipeline leverages occluded-unoccluded image pairs and pseudo-ground-truth views to teach the model structure-aware completion and view consistency. Without modifying the original architecture, we fully fine-tune the view synthesis model to jointly learn completion and multi-view generation. Additionally, we introduce the first benchmark for occlusion-aware reconstruction, covering diverse occlusion levels, object categories, and masking patterns, providing a standardized protocol for future evaluation.
format Preprint
id arxiv_https___arxiv_org_abs_2506_21544
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle DeOcc-1-to-3: 3D De-Occlusion from a Single Image via Self-Supervised Multi-View Diffusion
Qu, Yansong
Dai, Shaohui
Li, Xinyang
Wang, Yuze
Shen, You
Cao, Liujuan
Ji, Rongrong
Computer Vision and Pattern Recognition
Reconstructing 3D objects from a single image remains challenging, especially under real-world occlusions. While recent diffusion-based view synthesis models can generate consistent novel views from a single RGB image, they typically assume fully visible inputs and fail when parts of the object are occluded, resulting in degraded 3D reconstruction quality. We propose DeOcc-1-to-3, an end-to-end framework for occlusion-aware multi-view generation that synthesizes six structurally consistent novel views directly from a single occluded image, enabling reliable 3D reconstruction without prior inpainting or manual annotations. Our self-supervised training pipeline leverages occluded-unoccluded image pairs and pseudo-ground-truth views to teach the model structure-aware completion and view consistency. Without modifying the original architecture, we fully fine-tune the view synthesis model to jointly learn completion and multi-view generation. Additionally, we introduce the first benchmark for occlusion-aware reconstruction, covering diverse occlusion levels, object categories, and masking patterns, providing a standardized protocol for future evaluation.
title DeOcc-1-to-3: 3D De-Occlusion from a Single Image via Self-Supervised Multi-View Diffusion
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2506.21544