Dynamic Proximal Unrolling Network for Compressive Imaging

TL;DR

This paper introduces DPUNet, a versatile deep neural network that adaptively reconstructs images from compressive measurements across various modalities and sampling ratios without retraining.

Contribution

The paper proposes a dynamic proximal unrolling network that dynamically adjusts its parameters during inference, enabling multi-modal compressive imaging reconstruction with a single trained model.

Findings

DPUNet outperforms state-of-the-art methods across multiple imaging modalities.

It effectively handles varying sampling ratios and noise levels.

The dynamic proximal mapping enhances adaptability during inference.

Abstract

Compressive imaging aims to recover a latent image from under-sampled measurements, suffering from a serious ill-posed inverse problem. Recently, deep neural networks have been applied to this problem with superior results, owing to the learned advanced image priors. These approaches, however, require training separate models for different imaging modalities and sampling ratios, leading to overfitting to specific settings. In this paper, a dynamic proximal unrolling network (dubbed DPUNet) was proposed, which can handle a variety of measurement matrices via one single model without retraining. Specifically, DPUNet can exploit both the embedded observation model via gradient descent and imposed image priors by learned dynamic proximal operators, achieving joint reconstruction. A key component of DPUNet is a dynamic proximal mapping module, whose parameters can be dynamically adjusted at the inference stage and make it adapt to different imaging settings. Experimental results demonstrate that the proposed DPUNet can effectively handle multiple compressive imaging modalities under varying sampling ratios and noise levels via only one trained model, and outperform the state-of-the-art approaches.