arXiv cs.AI· Jinjing Zhao, Fangyun Wei, Yitong Wang, Xiuyu Wu, Yunuo Chen, Yang Yue, Sirui Zhang, Wenbo Wang, Hongyang Zhang, Dong Chen, Yan Lu, Chang Xu·· 9 天前AI 评分37
VibeEdit：用画布指令进行图像编辑
VibeEdit: Image Editing with Canvas Instructions
AI 导读
VibeEdit 提出一种画布指令图像编辑界面，用户直接在图上标注空间标记和简短备注来指定编辑位置与内容，无需单独文本提示词，即可完成对象添加、删除、替换、属性修改和移动。
正文 · 原文
Authors:Jinjing Zhao, Fangyun Wei, Yitong Wang, Xiuyu Wu, Yunuo Chen, Yang Yue, Sirui Zhang, Wenbo Wang, Hongyang Zhang, Dong Chen, Yan Lu, Chang Xu
Abstract:In text-guided image editing, describing the desired change is often straightforward, but identifying the intended object or region can be cumbersome, especially when several objects look alike. We introduce a new image editing interface that lets users place spatial marks and optional short notes directly on the image. Together, these annotations form a canvas instruction that specifies where to edit and what to change. Our editor, VibeEdit, follows these instructions to perform object addition, removal, replacement, attribute modification, and movement without a separate text prompt. We construct 1.55 million source-target edit pairs with object masks and structured edit descriptions, from which we render canvas instructions during training. We adapt Qwen-Image-Edit with layer-decoupled conditioning that separately encodes source images and canvas instructions for image editing. We train the model with region-weighted supervised fine-tuning, followed by rubric-guided reinforcement learning to improve edit completion, local edit quality, and preservation of unedited regions. We evaluate VibeEdit on an independently constructed, human-curated benchmark of 419 cases emphasizing target selection among similar objects. VibeEdit achieves a VLM rubric score of 79.9 and an outside-region PSNR of 32.8 dB, compared with 67.4 and 24.0 dB for FireRed, the highest-scoring text-instructed baseline in our evaluation.
| Comments: | Project page: this https URL |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.12229 [cs.CV] |
| (or arXiv:2610.12229v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.12229 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Jinjing Zhao [view email]
[v1]
Thu, 8 Oct 2026 16:15:46 UTC (21,370 KB)
来源：arXiv cs.AI · arxiv.org