SpatialEdit: Benchmarking Fine-Grained Image Spatial Editing

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Xiao, Yicheng, Zhang, Wenhu, Song, Lin, Chen, Yukang, Li, Wenbo, Jiang, Nan, Ren, Tianhe, Lin, Haokun, Huang, Wei, Huang, Haoyang, Li, Xiu, Duan, Nan, Qi, Xiaojuan
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911575744446464
author Xiao, Yicheng
Zhang, Wenhu
Song, Lin
Chen, Yukang
Li, Wenbo
Jiang, Nan
Ren, Tianhe
Lin, Haokun
Huang, Wei
Huang, Haoyang
Li, Xiu
Duan, Nan
Qi, Xiaojuan
author_facet Xiao, Yicheng
Zhang, Wenhu
Song, Lin
Chen, Yukang
Li, Wenbo
Jiang, Nan
Ren, Tianhe
Lin, Haokun
Huang, Wei
Huang, Haoyang
Li, Xiu
Duan, Nan
Qi, Xiaojuan
contents Image spatial editing performs geometry-driven transformations, allowing precise control over object layout and camera viewpoints. Current models are insufficient for fine-grained spatial manipulations, motivating a dedicated assessment suite. Our contributions are listed: (i) We introduce SpatialEdit-Bench, a complete benchmark that evaluates spatial editing by jointly measuring perceptual plausibility and geometric fidelity via viewpoint reconstruction and framing analysis. (ii) To address the data bottleneck for scalable training, we construct SpatialEdit-500k, a synthetic dataset generated with a controllable Blender pipeline that renders objects across diverse backgrounds and systematic camera trajectories, providing precise ground-truth transformations for both object- and camera-centric operations. (iii) Building on this data, we develop SpatialEdit-16B, a baseline model for fine-grained spatial editing. Our method achieves competitive performance on general editing while substantially outperforming prior methods on spatial manipulation tasks. All resources will be made public at https://github.com/EasonXiao-888/SpatialEdit.
format Preprint
id arxiv_https___arxiv_org_abs_2604_04911
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle SpatialEdit: Benchmarking Fine-Grained Image Spatial Editing
Xiao, Yicheng
Zhang, Wenhu
Song, Lin
Chen, Yukang
Li, Wenbo
Jiang, Nan
Ren, Tianhe
Lin, Haokun
Huang, Wei
Huang, Haoyang
Li, Xiu
Duan, Nan
Qi, Xiaojuan
Computer Vision and Pattern Recognition
Image spatial editing performs geometry-driven transformations, allowing precise control over object layout and camera viewpoints. Current models are insufficient for fine-grained spatial manipulations, motivating a dedicated assessment suite. Our contributions are listed: (i) We introduce SpatialEdit-Bench, a complete benchmark that evaluates spatial editing by jointly measuring perceptual plausibility and geometric fidelity via viewpoint reconstruction and framing analysis. (ii) To address the data bottleneck for scalable training, we construct SpatialEdit-500k, a synthetic dataset generated with a controllable Blender pipeline that renders objects across diverse backgrounds and systematic camera trajectories, providing precise ground-truth transformations for both object- and camera-centric operations. (iii) Building on this data, we develop SpatialEdit-16B, a baseline model for fine-grained spatial editing. Our method achieves competitive performance on general editing while substantially outperforming prior methods on spatial manipulation tasks. All resources will be made public at https://github.com/EasonXiao-888/SpatialEdit.
title SpatialEdit: Benchmarking Fine-Grained Image Spatial Editing
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2604.04911