用物理一致参数统一建模多种图像退化，实现零样本高效修复
Self-supervised Dynamic Heterogeneous Degradation Modeling for Unified Zero-Shot Image Restoration
- 将异质退化重参数为物理一致的紧凑参数，统一建模到单一分布
- 在潜在空间直接优化分布，提升修复质量并降低采样成本
- 动态调整扩散路径，避免陷入次优解，适合复杂退化场景
零样本图像修复可灵活应对多样退化而无需特定任务训练。现有方法多依赖堆叠层或预训练特征增强退化表达，却忽视了物理一致性先验。退化提示不足导致零样本扩散过程需大量训练和采样。此外，固定推理轨迹常在复杂退化下坍缩至次优解。我们发现异质退化可重参数化为一组物理连贯的最小参数，实现紧凑表征。基于此，提出统一物理零样本图像修复（UP-ZeroIR）框架，显式将异质退化建模为同质的全合一分布。该分布可在潜在空间直接优化，支持合理解探索与有效提示适配。同时引入动态质量精炼策略，自适应调整扩散轨迹，实现稳健全局最优收敛。大量实验表明，该方法在单类与混合退化上均达到领先性能。代码已开源：https://github.com/yangjinglyy/UP-ZeroIR
原文摘要 · Abstract (English)
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
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。