OARS：面向生成真实世界图像超分辨率的在线对齐
OARS: Process-Aware Online Alignment for Generative Real-World Image Super-Resolution
AI总结：
OARS通过过程感知的在线对齐框架，结合COMPASS奖励模型，在线优化图像超分辨率模型，实现感知与保真度的平衡，提升真实世界图像超分辨率性能。
摘要：
将生成真实世界图像超分辨率模型与人类视觉偏好对齐具有挑战性，由于感知与保真度的权衡以及多样且未知的退化。先前方法依赖于离线偏好优化和静态度量聚合，通常不可解释且在强条件下容易产生伪多样性。我们提出了OARS，一个基于COMPASS的基于大语言模型（MLLM）的奖励模型，通过联合建模保真度保持和感知增益，利用输入质量自适应的权衡策略来评估LR到SR的过渡。为训练COMPASS，我们整理了涵盖合成和真实退化的COMPASS-20K数据集，并引入了三阶段感知标注流程，产生校准且细粒度的训练标签。在COMPASS的指导下，OARS通过浅层LoRA优化进行渐进式在线对齐，从冷启动流匹配到全参考和最终无参考的强化学习。大量实验和用户研究证明了在保持保真度的同时，感知性能的一致提升，在Real-ISR基准上实现了最先进的性能。
英文摘要：
Aligning generative real-world image super-resolution models with human visual preference is challenging due to the perception--fidelity trade-off and diverse, unknown degradations. Prior approaches rely on offline preference optimization and static metric aggregation, which are often non-interpretable and prone to pseudo-diversity under strong conditioning. We propose OARS, a process-aware online alignment framework built on COMPASS, a MLLM-based reward that evaluates the LR to SR transition by jointly modeling fidelity preservation and perceptual gain with an input-quality-adaptive trade-off. To train COMPASS, we curate COMPASS-20K spanning synthetic and real degradations, and introduce a three-stage perceptual annotation pipeline that yields calibrated, fine-grained training labels. Guided by COMPASS, OARS performs progressive online alignment from cold-start flow matching to full-reference and finally reference-free RL via shallow LoRA optimization for on-policy exploration. Extensive experiments and user studies demonstrate consistent perceptual improvements while maintaining fidelity, achieving state-of-the-art performance on Real-ISR benchmarks.
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。
