Mapping a truncated optimization method into a deep neural network, deep
unfolding network (DUN) has attracted growing attention in compressive sensing
(CS) due to its good interpretability and high performance. Each stage in DUNs
corresponds to one iteration in optimization. By unders
本文提出了一种基于 DUN 框架下的 3D 卷积 - Transformer 混合(CTM)模块,该模块利用 Transformer 的 3D 有效可扩展关注模型充分学习时间和空间维度之间的相关性,并引入方差估计来表征重建过程中的高频信息,实验结果表明该模型在视频 SCI 重建方面取得了最好的表现(比此前的 SOTA 算法 PSNR 提高了 1.2dB)。