Generative real-world image super-resolution (Real-ISR) can synthesize visually convincing details from severely degraded low-resolution (LR) inputs, yet its stochastic sampling makes a critical failure mode hard to avoid: outputs may look sharp but be unfaithful to the LR evidence (semantic and structural hallucination), while such LR-anchored faithfulness is difficult to assess without HR ground truth. Preference-based reinforcement learning (RL) is a natural fit because each LR input yields a rollout group of candidates to compare. However, effective alignment in Real-ISR is hindered by (i) the lack of a degradation-robust LR-referenced faithfulness signal, and (ii) a rollout-group optimization bottleneck where naive multi-reward scalarization followed by normalization compresses objective-wise contrasts, causing advantage collapse and weakening the reward-weighted updates in DiffusionNFT-style forward fine-tuning. Moreover, (iii) limited coverage of real degradations restricts rollout diversity and preference signal quality. We propose LucidNFT, a multi-reward RL framework for flow-matching Real-ISR. LucidNFT introduces LucidConsistency, a degradation-robust semantic evaluator that makes LR-anchored faithfulness measurable and optimizable; a decoupled advantage normalization strategy that preserves objective-wise contrasts within each LR-conditioned rollout group before fusion, preventing advantage collapse; and LucidLR, a large-scale collection of real-world degraded images to support robust RL fine-tuning. Experiments show that LucidNFT consistently improves strong flow-based Real-ISR baselines, achieving better perceptual-faithfulness trade-offs with stable optimization dynamics across diverse real-world scenarios.

翻译：生成式真实世界图像超分辨率（Real-ISR）能从严重退化的低分辨率（LR）输入中合成视觉上令人信服的细节，但其随机采样特性导致一种关键失效模式难以避免：输出图像可能清晰却与LR证据不符（语义与结构幻觉），而缺乏高分辨率（HR）真值使得这种基于LR的保真度评估十分困难。基于偏好的强化学习（RL）天然适用于此场景，因为每个LR输入均可生成一组候选采样结果用于比较。然而，Real-ISR中的有效对齐面临三个障碍：（i）缺乏对退化具有鲁棒性的LR参考保真度信号；（ii）采样组优化瓶颈——简单采用多奖励标量化后归一化的方法会压缩目标间对比度，导致优势坍塌，削弱DiffusionNFT式前向微调中基于奖励加权的更新效果；以及（iii）真实退化的覆盖范围有限，限制了采样多样性与偏好信号质量。为此，我们提出LucidNFT——面向流匹配Real-ISR的多奖励强化学习框架。该框架包含三大贡献：LucidConsistency——一种对退化鲁棒的语义评估器，使基于LR锚定的保真度可量化、可优化；解耦优势归一化策略——在融合前保留每个LR条件采样组内的目标间对比度，避免优势坍塌；以及LucidLR——大规模真实退化图像数据集，支撑鲁棒的RL微调。实验表明，LucidNFT能持续改进基于流的Strong Real-ISR基线方法，在各类真实场景中实现更优的感知-保真度平衡，并保持稳定的优化动力学特性。