Abstract
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.
Abstract (translated)
生成型真实世界图像超分辨率（Real-ISR）能够从严重降级的低分辨率（LR）输入中合成视觉上令人信服的细节，但其随机采样使得一个关键失败模式难以避免：输出可能看起来很锐利但实际上却与LR证据不符（语义和结构错觉），而没有高分辨率（HR）真实情况作为参考，这种基于LR锚定的真实性很难评估。基于偏好的强化学习（RL）是一个自然的解决方案，因为每个低分辨率输入都会产生一组候选者来进行比较。然而，在Real-ISR中有效对齐受到以下问题的影响：(i) 缺乏一种抗降级且基于LR的真实度信号；以及(ii) 一个滚动组优化瓶颈，其中简单的多奖励标量化和随后的归一化压缩了每个目标之间的对比度，导致优势崩溃并减弱了DiffusionNFT式正向微调中的加权更新。此外，(iii) 真实降级范围有限也限制了滚动多样性及偏好信号质量。 为了解决这些问题，我们提出了LucidNFT，这是一个用于流匹配Real-ISR的多奖励RL框架。LucidNFT引入了一个新的方法——LucidConsistency，这是一种抗降级且基于语义的真实度评估器，使基于LR锚定的真实性变得可测量和优化；一种分离优势归一化策略，在融合之前保留每个基于低分辨率条件滚动组内的目标对比度，防止优势崩溃；以及LucidLR，一个大规模的包含真实世界降级图像的数据集，支持鲁棒RL微调。实验表明，LucidNFT能够持续改进强大的流型Real-ISR基线，并在各种现实场景中实现更好的感知真实性权衡和稳定的优化动态。 这个框架解决了现有方法面临的关键挑战，通过引入新的评估标准、优化策略以及大规模的真实数据集，从而提高了生成图像的质量与真实性。
URL
https://arxiv.org/abs/2603.05947
