提出迭代式视觉生成框架R^3，并构建600+专家标注数据集评测模型反思修正能力。
Benchmarking and Evolving Reason-Reflect-Rectify for Reflective Visual Generation
- 设计反思生成循环R^3，支持多轮迭代优化图像生成
- 在R^3-Bench上，新方法使反思评分提升12.0%，修正评分提升9.0%
- 适用于需高精度图像生成的科研与工业场景
文本到图像（T2I）模型和统一多模态模型（UMMs）在视觉生成方面取得了显著进展。然而，其依赖单次生成范式，难以处理需要迭代优化的复杂提示。为实现多轮反思式视觉生成（RVG），我们形式化了反思-反思-修正（R^3）循环作为核心框架，并引入R^3-Bench，一个包含600多个专家标注实例的基准测试，用于量化模型的迭代推理与修正能力。在R^3-Bench上的评估揭示了一个关键差距：尽管先进模型能识别生成错误，却无法生成可执行的修正指令。为弥合这一差距，我们提出R^3-Refiner，一种基于组相对策略优化（GRPO）和分层奖励机制（HRM）的双阶段框架，以更好对齐修正与反思推理。实验表明，R^3-Refiner在R^3-Bench上取得显著提升（反思判定得分+12.0%，修正得分+9.0%），且可无缝集成至多种多模态大模型中，提升T2I模型在GenEval++和T2I-CompBench上的生成质量。代码已开源：https://github.com/xiaomoguhz/R3-Bench。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。