Salvato in:
Dettagli Bibliografici
Autori principali: Cao, Xiao, Li, Yuze, Zhang, Youmin, Song, Jiayu, Yan, Cheng, Li, Wen, Duan, Lixin
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:https://arxiv.org/abs/2605.12399
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866918497857044480
author Cao, Xiao
Li, Yuze
Zhang, Youmin
Song, Jiayu
Yan, Cheng
Li, Wen
Duan, Lixin
author_facet Cao, Xiao
Li, Yuze
Zhang, Youmin
Song, Jiayu
Yan, Cheng
Li, Wen
Duan, Lixin
contents 3D Gaussian Splatting (3DGS) has emerged as a prominent paradigm for 3D reconstruction and novel view synthesis. However, it remains vulnerable to severe artifacts when trained under sparse-view constraints. While recent methods attempt to rectify artifacts in rendered views using image diffusion models, they typically rely on multi-view self-attention to retrieve information from reference images. We observe that this mechanism often fails when the rendered novel views output by 3DGS are heavily corrupted: damaged query features lead to erroneous cross-view retrieval, resulting in inconsistent rendering refinement. To address this, we propose GeoQuery, a geometry-guided diffusion framework that integrates generative priors with explicit geometric cues via a novel Geometry-guided Cross-view Attention (GCA) mechanism. First, by leveraging predicted depth maps and camera poses, we construct a geometry-induced correspondence field to sample reference features, forming a geometry-aligned proxy query that replaces the corrupted rendering features. Furthermore, we design a new cross-view feature aggregation pipeline, in which we restrict the cross-view attention to a local window around each proxy query to effectively retrieve useful features while suppressing spurious matches. GeoQuery can be seamlessly integrated into existing diffusion-based pipelines, enabling robust reconstruction even under extreme view sparsity. Extensive experiments on sparse-view novel view synthesis and rendering artifact removal demonstrate the effectiveness of our approach.
format Preprint
id arxiv_https___arxiv_org_abs_2605_12399
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle GeoQuery: Geometry-Query Diffusion for Sparse-View Reconstruction
Cao, Xiao
Li, Yuze
Zhang, Youmin
Song, Jiayu
Yan, Cheng
Li, Wen
Duan, Lixin
Computer Vision and Pattern Recognition
3D Gaussian Splatting (3DGS) has emerged as a prominent paradigm for 3D reconstruction and novel view synthesis. However, it remains vulnerable to severe artifacts when trained under sparse-view constraints. While recent methods attempt to rectify artifacts in rendered views using image diffusion models, they typically rely on multi-view self-attention to retrieve information from reference images. We observe that this mechanism often fails when the rendered novel views output by 3DGS are heavily corrupted: damaged query features lead to erroneous cross-view retrieval, resulting in inconsistent rendering refinement. To address this, we propose GeoQuery, a geometry-guided diffusion framework that integrates generative priors with explicit geometric cues via a novel Geometry-guided Cross-view Attention (GCA) mechanism. First, by leveraging predicted depth maps and camera poses, we construct a geometry-induced correspondence field to sample reference features, forming a geometry-aligned proxy query that replaces the corrupted rendering features. Furthermore, we design a new cross-view feature aggregation pipeline, in which we restrict the cross-view attention to a local window around each proxy query to effectively retrieve useful features while suppressing spurious matches. GeoQuery can be seamlessly integrated into existing diffusion-based pipelines, enabling robust reconstruction even under extreme view sparsity. Extensive experiments on sparse-view novel view synthesis and rendering artifact removal demonstrate the effectiveness of our approach.
title GeoQuery: Geometry-Query Diffusion for Sparse-View Reconstruction
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2605.12399