超越真实标签：利用图像质量先验进行真实世界图像复原
Beyond Ground-Truth: Leveraging Image Quality Priors for Real-World Image Restoration
利用预训练 NR-IQA 模型提取图像质量先验，作为条件信号引导复原网络预测高质量表示，可即插即用地增强现有复原架构；双分支码本与离散表示优化策略为质量感知复原提供了可复用的技术手段。
原文依据 we propose a novel framework, termed IQPIR, that introduces an Image Quality Prior (IQP)--extracted from pre-trained No-Reference Image Quality Assessment (NR-IQA) models--to guide the restoration process toward perceptually optimal outputs explicitly.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论真实世界图像复原旨在从不受控条件下捕获的退化低质量（LQ）输入中恢复高质量（HQ）图像。现有方法通常依赖真实标签（GT）监督，假设 GT 提供了完美的参考质量。然而，GT 中仍可能包含感知保真度不一致的图像，导致模型收敛到训练数据的平均质量水平，而非达到可获得的最高感知质量。为解决这些问题，我们提出了一种新颖的框架，称为 IQPIR，它引入图像质量先验（IQP）——从预训练的无参考图像质量评估（NR-IQA）模型中提取——以显式地引导复原过程朝向感知最优输出。我们的方法通过三个关键机制将 IQP 与学习到的码本先验协同集成：（1）质量条件 Transformer，其中 NR-IQA 导出的分数作为条件信号，引导预测表示朝向最大感知质量。该设计提供了即插即用的增强，与现有复原架构兼容且无需结构修改；（2）双分支码本结构，解耦通用特征和 HQ 特定特征，确保对通用结构信息和质量敏感属性的全面表示；（3）基于离散表示的质量优化策略，缓解连续潜变量空间中常见的过度优化效应。在真实世界图像复原上的大量实验表明，我们的方法不仅超越了前沿方法，而且可作为现有方法的通用质量引导增强策略。代码可在 https://github.com/fengyang1399-pixel/IQPIR 获取。
英文原文摘要 1823 字符
Real-world image restoration aims to restore high-quality (HQ) images from degraded low-quality (LQ) inputs captured under uncontrolled conditions. Existing methods typically depend on ground-truth (GT) supervision, assuming that GT provides perfect reference quality. However, GT can still contain images with inconsistent perceptual fidelity, causing models to converge to the average quality level of the training data rather than achieving the highest perceptual quality attainable. To address these problems, we propose a novel framework, termed IQPIR, that introduces an Image Quality Prior (IQP)--extracted from pre-trained No-Reference Image Quality Assessment (NR-IQA) models--to guide the restoration process toward perceptually optimal outputs explicitly. Our approach synergistically integrates IQP with a learned codebook prior through three key mechanisms:(1) a quality-conditioned Transformer, where NR-IQA-derived scores serve as conditioning signals to steer the predicted representation toward maximal perceptual quality. This design provides a plug-and-play enhancement compatible with existing restoration architectures without structural modification; and(2) a dual-branch codebook structure, which disentangles common and HQ-specific features, ensuring a comprehensive representation of both generic structural information and quality-sensitive attributes; and (3) a discrete representation-based quality optimization strategy, which mitigates over-optimization effects commonly observed in continuous latent spaces. Extensive experiments on real-world image restoration demonstrate that our method not only surpasses cutting-edge methods but also serves as a generalizable quality-guided enhancement strategy for existing methods. The code is available at https://github.com/fengyang1399-pixel/IQPIR.