arXiv cs.AI· Jun He, Canmiao Fu, Ying Zhang, Zhipeng Huang, Chenyu Mu, Yazhen Xie, Chen Li, Jing Lyu, Weijia Li, Xiang Zhang·· 4 小时前AI 评分40
WeGenEditBench：从评测到进化的图像编辑基准与 WeGenEvoStudio 框架
From Evaluation to Evolution: Benchmarking Image Editing and Evolving Image Models
AI 导读
一项新研究推出图像编辑基准 WeGenEditBench，包含 2,740 个评测实例和 148 种任务类型，覆盖单图、文本密集和多参考编辑。在 15 个受测配置中，GPT-Image-2 以 4.076 的 Overall 分领先单图编辑，Seedream-5.0-Pro 以 4.291 领先文本密集编辑。
正文 · 原文
Abstract:Real-world image editing requires coordinating multiple visual transformations with content preservation and reference constraints, yet aggregate scores can readily obscure unmet local requirements. Effective evaluation must therefore identify local failures and translate them into priorities for subsequent model improvement. We introduce WeGenEditBench, comprising 2,740 evaluation instances and 148 task types across single-image, text-intensive, and multi-reference editing. Pre-generated checklists shared across models decompose each request into atomic requirements; item-level deductions and visual evidence are established first and then summarized into scores for edit execution, preservation, visual quality, and task-specific correctness. To support the experience-acquisition component of recursive self-improvement (RSI), we develop WeGenEvoStudio, a unified multi-agent harness for RSI-oriented learning that integrates model-conditioned failure diagnosis, retrieval-grounded data construction, supervised adaptation, and re-evaluation. Among 15 evaluated configurations, GPT-Image-2 leads single-image editing with an Overall score of 4.076, while Seedream-5.0-Pro leads text-intensive editing at 4.291. Evolution experiments across four generation and editing benchmarks demonstrate substantial targeted gains: on their report-selected Top-5 single-image targets, FLUX.2 [dev] improves Overall from 1.083 to 3.517, while Qwen-Image-Edit-2511 gains 13.2%. Together, our benchmark and harness provide a foundation for developing more reliable image models through interpretable evaluation and feedback-driven learning.
| Comments: | 58 pages, 42 figures |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.41210 [cs.CV] |
| (or arXiv:2610.41210v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.41210 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Jun He [view email]
[v1]
Fri, 30 Oct 2026 03:42:15 UTC (44,099 KB)
来源：arXiv cs.AI · arxiv.org