面向统一零样本图像复原的自监督动态异质退化建模
Self-supervised Dynamic Heterogeneous Degradation Modeling for Unified Zero-Shot Image Restoration
把异质退化重参数化为少量物理一致的参数，并在潜空间中直接优化该分布，为统一零样本复原提供了一种紧凑的退化提示词构造方式；其动态质量精炼策略可迁移到其他扩散采样轨迹的自适应调整中。
原文依据 heterogeneous degradations can be reparameterized into a minimal set of physically coherent parameters for compact representation
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论零样本图像复原提供了一种无需针对特定任务训练即可处理多样退化的灵活方式。然而，现有方法通常依赖堆叠层或预训练特征来增强退化表达，却忽视了物理一致的先验。退化提示词不足，给零样本扩散带来了沉重的训练负担和高昂的采样成本。此外，固定的推理轨迹在复杂损坏下往往收敛到次优解。我们观察到，异质退化可以被重参数化为最小的一组物理一致参数，从而实现紧凑表示。基于这一洞见，我们首先提出了统一物理零样本图像复原（UP-ZeroIR）框架，将异质退化显式建模为同质的统一分布。该分布可以直接在潜变量空间中优化，从而实现有原则的解探索和有效的提示词自适应。此外，我们引入了动态质量精炼策略，自适应地调整扩散轨迹，以实现稳健的全局最优收敛。大量实验表明，我们的方法在单一退化和混合退化上均达到了最优性能。代码见 https://github.com/yangjinglyy/UP-ZeroIR。
英文原文摘要 1346 字符
Zero-shot image restoration provides a flexible way to handle diverse degradations without task-specific training. However, existing methods typically rely on stacked layers or pre-trained features to enhance degradation expression, while overlooking physically consistent priors. The insufficient degradation prompts impose the heavy training burden and high sampling costs during zero-shot diffusion. Moreover, the fixed inference trajectory often collapses to suboptimal solutions under complex corruptions. We observe that heterogeneous degradations can be reparameterized into a minimal set of physically coherent parameters for compact representation. Based on this insight, we first propose a unified physical zero-shot image restoration (UP-ZeroIR) framework that explicitly models heterogeneous degradations into a homogeneous all-in-one distribution. The distribution can be optimized directly in the latent space, enabling principled solution exploration and effective prompt adaptation. Besides, we introduce a dynamic quality-refinement strategy that adaptively adjusts the diffusion trajectory for robust globally optimal convergence. Extensive experiments demonstrate that our method achieves state-of-the-art performance across both single and mixed degradations. Our code is available at https://github.com/yangjinglyy/UP-ZeroIR.