Model Adaptation for Inverse Problems in Imaging
Model Adaptation for Inverse Problems in Imaging
复制标题
DOI：
10.1109/tci.2021.3094714
复制
发表时间：
2021-01-01
影响因子：
5.4
通讯作者：
Willett, Rebecca
中科院分区：
文献类型：
--
作者：
Gilton, Davis;Ongie, Gregory;Willett, Rebecca
深度神经网络已成功应用于计算成像中出现的各种逆问题。这些网络通常使用描述待反演的测量过程的前向模型进行训练，该模型通常直接并入网络本身。然而，这些方法对前向模型的变化很敏感：如果在测试时前向模型与网络训练的模型有差异（即使是轻微的），重建性能可能会大幅下降。给定一个经过训练的网络，用一个已知的正向模型来解决初始逆问题，我们提出了两个新的程序，使网络适应正向模型的变化，即使没有充分的知识的变化。我们的方法不需要访问更多的标记数据（即，地面实况图像）。我们证明了这些简单的模型自适应方法在各种逆问题中取得了经验上的成功，包括去模糊，超分辨率和磁共振成像中的欠采样图像重建。
Deep neural networks have been applied successfully to a wide variety of inverse problems arising in computational imaging. These networks are typically trained using a forward model that describes the measurement process to be inverted, which is often incorporated directly into the network itself. However, these approaches are sensitive to changes in the forward model: if at test time the forward model varies (even slightly) from the one the network was trained for, the reconstruction performance can degrade substantially. Given a network trained to solve an initial inverse problem with a known forward model, we propose two novel procedures that adapt the network to a change in the forward model, even without full knowledge of the change. Our approaches do not require access to more labeled data (i.e., ground truth images). We show these simple model adaptation approaches achieve empirical success in a variety of inverse problems, including deblurring, super-resolution, and undersampled image reconstruction in magnetic resonance imaging.