ISTOF-Net-main

ISTOF-Net: An ISTA-Based Deep Unfolding Network with Optimized Feature Aggregation Architecture for Image Compressed SensingCompressed Sensing (CS) provides an efficient approach for data acquisition and reconstruction, has become a prevalent technique in image processing. Deep unfolding networks integrate deep neural networks with optimization algorithms, emerging as a prevalent approach for image CS by enhancing interpretability and reconstruction performance. However, most existing deep unfolding networks suffer from insufficient interaction and utilization of feature information during the iterative process. To address this limitation and achieve the goal of fusing complementary information to form a richer and more fine feature representation, we propose an Iterative Shrinking-Threshold Algorithm (ISTA)-based deep unfolding network with optimized feature aggregation architecture for high-quality image CS, dubbed ISTOF-Net.

ISTOF-Net：一种基于迭代收缩阈值算法（Iterative Shrinking-Threshold Algorithm，ISTA）且优化特征聚合架构的图像压缩感知深度展开网络。压缩感知（Compressed Sensing，CS）为数据采集与重建提供了高效途径，现已成为图像处理领域的主流技术。深度展开网络将深度神经网络与优化算法相结合，通过提升模型可解释性与重建性能，成为图像压缩感知领域的主流方案。然而，现有多数深度展开网络在迭代过程中存在特征信息交互不充分、利用不足的问题。为解决这一局限，并实现融合互补信息以构建更丰富、更精细的特征表示的目标，本文提出一种基于迭代收缩阈值算法（ISTA）且优化特征聚合架构的高质量图像压缩感知深度展开网络，命名为ISTOF-Net。