Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2512.02790 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909940338130944 |
|---|---|
| author | Ye, Keming Huang, Zhipeng Fu, Canmiao Liu, Qingyang Cai, Jiani Lv, Zheqi Li, Chen Lyu, Jing Zhao, Zhou Zhang, Shengyu |
| author_facet | Ye, Keming Huang, Zhipeng Fu, Canmiao Liu, Qingyang Cai, Jiani Lv, Zheqi Li, Chen Lyu, Jing Zhao, Zhou Zhang, Shengyu |
| contents | With the rapid advances of powerful multimodal models such as GPT-4o, Nano Banana, and Seedream 4.0 in Image Editing, the performance gap between closed-source and open-source models is widening, primarily due to the scarcity of large-scale, high-quality training data and comprehensive benchmarks capable of diagnosing model weaknesses across diverse editing behaviors. Existing data construction methods face a scale-quality trade-off: human annotations are high-quality but not scalable, while automated pipelines suffer from error propagation and noise. To address this, we introduce a lightweight data pipeline that replaces multi-toolchains with an end-to-end model and a unified post-verification stage. For scalable quality control, we train a 7B dual-task expert model, \textbf{Qwen-Verify}, for efficient failure detection and instruction recaptioning. This pipeline yields \textbf{UnicEdit-10M}, a 10M-scale dataset spanning diverse basic and complex editing tasks. We also propose \textbf{UnicBench}, a general benchmark that extends beyond basic edits to explicitly assess spatial and knowledge-driven reasoning. To enable fine-grained diagnosis, we introduce novel metrics, including \textit{Non-edit Consistency} and \textit{Reasoning Accuracy}. Our analysis of mainstream models on UnicBench reveals their limitations and provides clear directions for future research. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_02790 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | UnicEdit-10M: A Dataset and Benchmark Breaking the Scale-Quality Barrier via Unified Verification for Reasoning-Enriched Edits Ye, Keming Huang, Zhipeng Fu, Canmiao Liu, Qingyang Cai, Jiani Lv, Zheqi Li, Chen Lyu, Jing Zhao, Zhou Zhang, Shengyu Computer Vision and Pattern Recognition With the rapid advances of powerful multimodal models such as GPT-4o, Nano Banana, and Seedream 4.0 in Image Editing, the performance gap between closed-source and open-source models is widening, primarily due to the scarcity of large-scale, high-quality training data and comprehensive benchmarks capable of diagnosing model weaknesses across diverse editing behaviors. Existing data construction methods face a scale-quality trade-off: human annotations are high-quality but not scalable, while automated pipelines suffer from error propagation and noise. To address this, we introduce a lightweight data pipeline that replaces multi-toolchains with an end-to-end model and a unified post-verification stage. For scalable quality control, we train a 7B dual-task expert model, \textbf{Qwen-Verify}, for efficient failure detection and instruction recaptioning. This pipeline yields \textbf{UnicEdit-10M}, a 10M-scale dataset spanning diverse basic and complex editing tasks. We also propose \textbf{UnicBench}, a general benchmark that extends beyond basic edits to explicitly assess spatial and knowledge-driven reasoning. To enable fine-grained diagnosis, we introduce novel metrics, including \textit{Non-edit Consistency} and \textit{Reasoning Accuracy}. Our analysis of mainstream models on UnicBench reveals their limitations and provides clear directions for future research. |
| title | UnicEdit-10M: A Dataset and Benchmark Breaking the Scale-Quality Barrier via Unified Verification for Reasoning-Enriched Edits |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2512.02790 |