arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2603.22228cs.CVcs.AI
收藏

SpatialReward: 可验证的空间奖励建模以实现文本到图像生成中的细粒度空间一致性

SpatialReward: Verifiable Spatial Reward Modeling for Fine-Grained Spatial Consistency in Text-to-Image Generation

Zhejiang University(浙江大学);Alibaba Group(阿里巴巴集团);Fudan University(复旦大学)

Sashuai Zhou, Qiang Zhou, Junpeng Ma, Yue Cao, Ruofan Hu, Ziang Zhang, Xiaoda Yang, Zhibin Wang, Jun Song, Cheng Yu, Bo Zheng, Zhou Zhao

AI总结:

本文提出SpatialReward,一种专门评估生成图像中空间布局的可验证奖励模型,通过多阶段流程提升空间一致性与生成质量,实验表明其能更贴近人类判断。

摘要:

最近基于强化学习(RL)的文本到图像(T2I)生成在奖励模型方面取得了进展,这些模型评估语义对齐和视觉质量。然而,大多数现有奖励模型对细粒度空间关系关注有限,常生成整体看似合理但存在物体位置不准确的图像。本文提出SpatialReward,一种专门评估生成图像中空间布局的可验证奖励模型。SpatialReward采用多阶段流程:Prompt Decomposer从自由形式提示中提取实体、属性和空间元数据;专家检测器提供准确的视觉定位和属性;视觉语言模型通过基于事实推理的观察评估复杂空间关系。为更全面评估生成图像中的空间关系,我们引入SpatRelBench基准,涵盖对象属性、方向、对象间关系和渲染文本位置。实验表明,将SpatialReward纳入RL训练可一致提升空间一致性和整体生成质量,结果更贴近人类判断。这些发现表明,可验证奖励模型在文本到图像生成模型中具有较大潜力,能够实现更准确和可控的优化。

英文摘要:

Recent advances in text-to-image (T2I) generation via reinforcement learning (RL) have benefited from reward models that assess semantic alignment and visual quality. However, most existing reward models pay limited attention to fine-grained spatial relationships, often producing images that appear plausible overall yet contain inaccuracies in object positioning. In this work, we present \textbf{SpatialReward}, a verifiable reward model explicitly designed to evaluate spatial layouts in generated images. SpatialReward adopts a multi-stage pipeline: a \emph{Prompt Decomposer} extracts entities, attributes, and spatial metadata from free-form prompts; expert detectors provide accurate visual grounding of object positions and attributes; and a vision-language model applies chain-of-thought reasoning over grounded observations to assess complex spatial relations that are challenging for rule-based methods. To more comprehensively evaluate spatial relationships in generated images, we introduce \textbf{SpatRelBench}, a benchmark covering object attributes, orientation, inter-object relations, and rendered text placement. Experiments on Stable Diffusion and FLUX show that incorporating SpatialReward into RL training consistently improves spatial consistency and overall generation quality, with results aligned more closely to human judgments. These findings indicate that verifiable reward models hold considerable potential for enabling more accurate and controllable optimization in text-to-image generation models.

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。

↑