arXiv cs.AI· Ching-Heng Cheng, Chia-Ming Lee, Ming-Ching Chang, Xin Li, Chih-Chung Hsu·· 2 小时前AI 评分37
FlashReFocus:秒级交互式单图像重新对焦
FlashReFocus: Interactive Image Refocusing in Seconds
AI 导读
研究者提出 FlashReFocus,每张照片只需运行一次扩散模型,后续编辑由轻量网络渲染,实现单张照片的交互式重新对焦。系统用单步扩散去模糊器恢复细节,渲染器通过显式虚化图和 defocus-bias attention 支持tilt-shift 等效果,并发布包含 89 组相机-镜头对的新基准 3CReal。
正文 · 原文
Abstract:Single-image refocusing changes the focal plane and depth of field of a photograph after capture. Users compare focal choices on one photograph, so content stability across edits matters alongside quality and speed. Recent generative methods rerun large diffusion models for every edit, even when they first restore an all-in-focus image, so content can change inconsistently between edits. Yet only this restoration is highly ill-posed and benefits from a generative prior, whereas rendering blur is largely determined by depth and the focal plane. We therefore propose FlashReFocus, which runs diffusion once per photograph and renders each edit with a lightweight network. A one-step diffusion deblurrer with a decoder refiner recovers fine detail efficiently while balancing fidelity and perceptual quality. The renderer uses an explicit defocus map for control and defocus-bias attention to promote information exchange between tokens with similar blur magnitudes. Because every edit reuses the same reconstruction, in-focus content stays consistent across edits, and editing the map enables tilt-shift and two-plane effects. We further collect 3CReal, a benchmark of 89 pairs from three camera-lens systems, to evaluate generalization to unseen optics. For zero-shot bokeh rendering on EBB! Val176 and 3CReal, FlashReFocus leads the evaluated methods in fidelity and perceptual quality. At 1536x1536, a complete refocus takes 2.45 s and each subsequent edit 0.46 s on one RTX PRO 6000.
| Comments: | 38 pages (8 main + appendix), 47 figures, 16 tables. Project page: this https URL Code: this https URL |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Image and Video Processing (eess.IV) |
| Cite as: | arXiv:2610.23297 [cs.CV] |
| (or arXiv:2610.23297v2 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.23297 arXiv-issued DOI via DataCite |
Submission history
From: Chia-Ming Lee [view email]
[v1]
Fri, 16 Oct 2026 11:04:18 UTC (33,812 KB)
[v2]
Wed, 18 Nov 2026 11:54:21 UTC (34,688 KB)
来源:arXiv cs.AI · arxiv.org