无需配对数据，用混合去噪正则实现测量只的图像重建
HyDRA: Hybrid Denoising Regularization for Measurement-Only DEQ Training
- 结合测量一致性与自适应去噪正则，仅用测量值训练
- 在稀疏视角CT上达到媲美监督方法的重建质量
- 适合无真实标签数据的医学成像场景
求解形如 \\(\mathbf{A} \mathbf{x} = \mathbf{y}\\) 的图像重建问题仍具挑战性，因问题不适定且缺乏大规模有监督数据集。深度平衡（DEQ）模型虽表现良好，但通常需配对数据 \\((\mathbf{x},\mathbf{y})\\)。在多数实际场景中，仅有测量值 \\(\\mathbf{y}\\) 可用。本文提出 HyDRA（混合去噪正则化适配），一种仅依赖测量值的 DEQ 训练框架，结合测量一致性与自适应去噪正则项，并引入数据驱动的早停机制。在稀疏视角计算机断层扫描（CT）任务上的实验表明，该方法可实现具有竞争力的重建质量并支持快速推理。
原文摘要 · Abstract (English)
Solving image reconstruction problems of the form \(\mathbf{A} \mathbf{x} = \mathbf{y}\) remains challenging due to ill-posedness and the lack of large-scale supervised datasets. Deep Equilibrium (DEQ) models have been used successfully but typically require supervised pairs \((\mathbf{x},\mathbf{y})\). In many practical settings, only measurements \(\mathbf{y}\) are available. We introduce HyDRA (Hybrid Denoising Regularization Adaptation), a measurement-only framework for DEQ training that combines measurement consistency with an adaptive denoising regularization term, together with a data-driven early stopping criterion. Experiments on sparse-view CT demonstrate competitive reconstruction quality and fast inference.
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。