基于推理的图像风格化代理规划 via 离线强化学习
Agentic Planning with Reasoning for Image Styling via Offline RL
AI总结：
本文提出基于推理的代理规划框架，通过结构化规划和工具链提升图像风格化效果，采用合成数据生成和离线强化学习方法，验证在视觉质量和指令遵循上的优越性。
摘要：
直接基于提示的编辑在复杂变换上往往失效，因为模糊和主观的提示通常需要对图像中应更改的部分有细致的理解。我们的核心直觉是，利用组合图像编辑工具而不是直接提示，可以借助结构化的代理级规划和明确的推理，从而获得更好的结果。这种结构化的规划框架使在训练后通过质量评分轨迹进行离线强化学习以提高性能成为可能。我们提出了一种基于工具的代理强化学习后训练框架，通过结构化的规划和推理链来解决这个问题。我们的关键贡献包括：(1) 一种基于工具的代理规划方法，结合了组合库中的正交基本变换、结构化的上下文表示和明确的每步推理，将复杂的风格化分解为可解释的工具序列。(2) 一个合成数据生成管道，产生三个大规模数据集（每个约为10,000条轨迹），包含推理链、计划和质量评分，因为现有的数据集没有提供这样的监督。(3) 离线强化学习训练方法，用于学习具有推理能力的规划器，作为我们的核心算法贡献，这些方法在视觉质量和指令遵循方面始终优于Edit-Only基线。(4) 在4B和8B参数Qwen3-VL模型上的全面评估，显示我们的方法在大多数组合任务中优于其他基线，经由人类评估验证。
英文摘要：
Direct prompt-based editing often fails on complex transformations because vague and subjective prompts often require nuanced understanding of what should be changed in the image. Our core intuition is that leveraging compositional image editing tools rather than direct prompting profits from structured agent-level planning with explicit reasoning, leading to better results. This structured planning framework enables efficient offline RL post-training on quality-scored trajectories to improve performance. We present a tool-based agentic RL post-training framework that addresses this through structured planning with chain-of-thought reasoning. Our key contributions include: (1) A tool-based agentic planning methodology that combines a compositional library of orthogonal primitive transformations, structured context representation, and explicit per-step reasoning to decompose complex styling into interpretable tool sequences. (2) A synthetic data generation pipeline producing three large-scale datasets (each $\sim$10K trajectories) with reasoning chains, plans, and quality scores, as no existing datasets provide such supervision. Our datasets and code are publicly available at the HuggingFace repository. (3) Offline RL training methods for learning planners with reasoning as our core algorithmic contributions, which consistently improve over the Edit-Only baseline in visual quality and instruction following. (4) Comprehensive evaluation across 4B and 8B parameter Qwen3-VL models showing that our methods outperform other baselines in the majority of compositional tasks, validated by human evaluations.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
