Benchmarking and Evolving Reason-Reflect-Rectify for Reflective Visual Generation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910236179169280 |
|---|---|
| author | Wang, Junjie Lou, Xinghua Li, Jason Tian, Ye Chen, Keyu Li, Yulin Kang, Bin Mai, Jacky Li, Yanwei Tian, Zhuotao Nie, Liqiang |
| author_facet | Wang, Junjie Lou, Xinghua Li, Jason Tian, Ye Chen, Keyu Li, Yulin Kang, Bin Mai, Jacky Li, Yanwei Tian, Zhuotao Nie, Liqiang |
| contents | Text-to-Image (T2I) models and Unified Multimodal Models (UMMs) have achieved remarkable progress in visual generation. However, their reliance on a single-pass generation paradigm limits their ability to handle complex prompts requiring iterative refinement. To enable multi-round Reflective Visual Generation (RVG), we formalize the Reason-Reflect-Rectify (R^3) loop as a core framework and introduce R^3-Bench, a benchmark of over 600 expert-annotated instances that quantifies iterative reasoning and rectification capabilities. Evaluation on R^3-Bench reveals a critical gap: while state-of-the-art models can identify generation errors, they fail to generate actionable rectification instructions. To bridge this gap, we propose R^3-Refiner, a dual-stage framework leveraging Group Relative Policy Optimization (GRPO) and a Hierarchical Reward Mechanism (HRM) to better align rectification with reflective reasoning. Experiments show that R^3-Refiner achieves significant improvements on R^3-Bench (+12.0% in Reflective Verdict Score, +9.0% in Rectification Score), and can be seamlessly integrated with various MLLMs to enhance the generation quality of different T2I models on GenEval++ and T2I-CompBench. Code is available at https://github.com/xiaomoguhz/R3-Bench. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_19639 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Benchmarking and Evolving Reason-Reflect-Rectify for Reflective Visual Generation Wang, Junjie Lou, Xinghua Li, Jason Tian, Ye Chen, Keyu Li, Yulin Kang, Bin Mai, Jacky Li, Yanwei Tian, Zhuotao Nie, Liqiang Computer Vision and Pattern Recognition Text-to-Image (T2I) models and Unified Multimodal Models (UMMs) have achieved remarkable progress in visual generation. However, their reliance on a single-pass generation paradigm limits their ability to handle complex prompts requiring iterative refinement. To enable multi-round Reflective Visual Generation (RVG), we formalize the Reason-Reflect-Rectify (R^3) loop as a core framework and introduce R^3-Bench, a benchmark of over 600 expert-annotated instances that quantifies iterative reasoning and rectification capabilities. Evaluation on R^3-Bench reveals a critical gap: while state-of-the-art models can identify generation errors, they fail to generate actionable rectification instructions. To bridge this gap, we propose R^3-Refiner, a dual-stage framework leveraging Group Relative Policy Optimization (GRPO) and a Hierarchical Reward Mechanism (HRM) to better align rectification with reflective reasoning. Experiments show that R^3-Refiner achieves significant improvements on R^3-Bench (+12.0% in Reflective Verdict Score, +9.0% in Rectification Score), and can be seamlessly integrated with various MLLMs to enhance the generation quality of different T2I models on GenEval++ and T2I-CompBench. Code is available at https://github.com/xiaomoguhz/R3-Bench. |
| title | Benchmarking and Evolving Reason-Reflect-Rectify for Reflective Visual Generation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2605.19639 |