Skip to main content
QUICK REVIEW

[论文解读] Benchmarking Uncertainty Quantification of Plug-and-Play Diffusion Priors for Inverse Problems Solving

Xiaoyu Qiu, Taewon Yang|arXiv (Cornell University)|Feb 4, 2026
Probabilistic and Robust Engineering Design被引用 0
一句话总结

论文基准测试了用于逆问题的可插件化扩散先验求解器(PnPDP)在不确定性量化(UQ)方面的表现,显示类似的重建质量可能隐藏截然不同的后验不确定性,并提出一个基于UQ的分类法与诊断框架。

ABSTRACT

Plug-and-play diffusion priors (PnPDP) have become a powerful paradigm for solving inverse problems in scientific and engineering domains. Yet, current evaluations of reconstruction quality emphasize point-estimate accuracy metrics on a single sample, which do not reflect the stochastic nature of PnPDP solvers and the intrinsic uncertainty of inverse problems, critical for scientific tasks. This creates a fundamental mismatch: in inverse problems, the desired output is typically a posterior distribution and most PnPDP solvers induce a distribution over reconstructions, but existing benchmarks only evaluate a single reconstruction, ignoring distributional characterization such as uncertainty. To address this gap, we conduct a systematic study to benchmark the uncertainty quantification (UQ) of existing diffusion inverse solvers. Specifically, we design a rigorous toy model simulation to evaluate the uncertainty behavior of various PnPDP solvers, and propose a UQ-driven categorization. Through extensive experiments on toy simulations and diverse real-world scientific inverse problems, we observe uncertainty behaviors consistent with our taxonomy and theoretical justification, providing new insights for evaluating and understanding the uncertainty for PnPDPs.

研究动机与目标

  • 在不适定逆问题中,激发对PnPDP求解器进行不确定性感知评估的需求。
  • 基于其近似贝叶斯后验的能力,提出一个以UQ驱动的PnPDP方法分类。
  • 开发包含已知真值后验的 toy 模型诊断工具,用以量化不确定性的校准并比较方法。
  • 在真实世界的科学逆问题中展示不确定性行为的一致性。
  • 为评估和理解扩散型逆求解器中的不确定性提供建议。

提出的方法

  • 定义后验目标 p(x|y) 并区分后验目标、启发式和类似MAP的PnPDP求解器。
  • 引入来自重复采样的经验后验方差,作为求解器诱发不确定性的代理指标。
  • 设计具有已知真实后验的toy实验证,以校准 AU 与 EU 并验证UQ指标。
  • 在多种PnPDP方法下评估真实数据的逆问题(线性散射、稀疏采样MRI、稀疏投影CT)。
  • 提供一个将方法与其后验目标能力及理论保证联系起来的基于UQ的分类法。
Figure 1 : Illustration of the Accuracy Trap phenomenon and three types of uncertainty behaviors. Blue contours show the ground-truth posterior $p(x\mid y)$ , which can be multi-modes. The red star denotes the ground-truth $x^{*}$ ; $\hat{x}_{1},\hat{x}_{3}$ are posterior-plausible reconstructions,
Figure 1 : Illustration of the Accuracy Trap phenomenon and three types of uncertainty behaviors. Blue contours show the ground-truth posterior $p(x\mid y)$ , which can be multi-modes. The red star denotes the ground-truth $x^{*}$ ; $\hat{x}_{1},\hat{x}_{3}$ are posterior-plausible reconstructions,

实验结果

研究问题

  • RQ1在不适定的前向模型下,随机的PnPDP求解器是否能够恢复后验 p(x|y) 及其不确定性?
  • RQ2不同的PnPDP求解器在经过校准的不确定性方面相比如何,而不仅仅是重建精度?
  • RQ3基于UQ的分类法是否能反映 toy 与真实数据任务中的不确定性行为?
  • RQ4测量稀疏性或分布外数据如何影响求解器的不确定性?
  • RQ5后验目标求解器的渐近保证与真实实现之间存在哪些局限性和实际差距?

主要发现

  • PnPDP求解器的不确定性行为差异可能很大,即使重建精度相近。
  • 后验目标求解器(如 MCG-Diff、FPS-SMC、PnPDM)在 toy 与部分真实数据测试中显示出经过校准的不确定性,但在实际中仍可能存在偏差或退化。
  • 类似MAP的求解器(如 REDDiff)给出接近零的方差,与点估计一致。
  • 启发式求解器展现出多样化的不确定性模式,在某些前向模型下可能无法正确校准不确定性。
  • 不确定性通常随测量稀疏化而增加,精度与不确定性是互补的评估维度。
  • 基于UQ的分类法与观察到的行为保持一致,并对算法结构分类提供了补充。
Figure 2 : Similar reconstruction with distinct uncertainty. Comparison of PnPDP solvers on linear inverse scattering reconstruction with $K=100$ times reconstruction on each solver. Top two Row: These methods produce similar reconstruction quality (in PSNR). Bottom Row: The pixel-wise variance maps
Figure 2 : Similar reconstruction with distinct uncertainty. Comparison of PnPDP solvers on linear inverse scattering reconstruction with $K=100$ times reconstruction on each solver. Top two Row: These methods produce similar reconstruction quality (in PSNR). Bottom Row: The pixel-wise variance maps

更好的研究,从现在开始

从阅读论文到最终审阅,大幅缩短您的研究时间。

无需绑定信用卡

本解读由 AI 生成,并经人工编辑审核。