用锚定机制稳定深度图像重建，防止迭代中突然崩溃。
Stabilizing Deep Reconstruction Operators with Contractive Anchoring
- 引入可收缩算子作为稳定锚点，动态控制重建过程
- 实验显示多任务、多噪声下均无崩溃，性能更可靠
- 无需重训练，适配各类预训练去噪器和重建算法
预训练的深度去噪器可通过插件式（PnP）和通过去噪正则化（RED）算法解决多种基于模型的图像重建任务，而无需为每个任务重新训练。这些去噪器仅针对单步去噪进行训练。在迭代重建过程中将其作为正则化器使用可能导致不稳定。常见失效模式为峰值与坍塌行为：早期迭代中如PSNR等指标上升，随后突然下降，使算法在实际中不可靠。本文提出一种数据驱动的稳定框架，（i）通过局部量形式化任意重建算子的不稳定性，（ii）通过自适应正则化该量防止坍塌，且无需对给定预训练网络进行重训练或修改。核心思想是利用一个具有收缩特性的算子作为稳定迭代的锚点，防止重建崩溃。我们进一步引入一组可训练的轻量级收缩算子，作为强锚点同时保持高效性。在多种邻近算法、去噪器架构、噪声水平和成像任务上的大量实验表明，该方法实现了持续稳定的重建，避免了坍塌，并显著提升了PnP与RED重建的可靠性。
原文摘要 · Abstract (English)
Pretrained deep denoisers can be used to solve a wide range of model-based image reconstruction tasks via Plug-and-Play (PnP) and Regularization-by-Denoising (RED) algorithms, without retraining per task. These denoisers are trained only for single-step denoising. Using them as Image Reconstruction (IR) regularizers in an iterative process can destabilize reconstruction. A common failure mode is the peak-and-collapse behaviour: metrics such as PSNR improve for early iterations and then abruptly degrade, making these algorithms unreliable in practice. We propose a data-driven stabilization framework that (i) formalizes this instability of any IR operator through a local quantity and (ii) prevents collapse by regularizing this quantity adaptively, requiring no retraining or modification of the given pretrained network. Our key idea is to control the potentially unstable IR operator with a contractive operator whose stable iterates act as an anchor and prevent collapse. We further introduce an efficient family of trainable contractive operators that serve as strong anchors while remaining lightweight. Extensive experiments across proximal algorithms, denoiser architectures, noise levels, and imaging tasks show consistent, collapse-free performance and improved reliability of PnP and RED reconstruction.
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。