深度BCD-Net：使用相同编码-解码CNN结构的迭代图像恢复
Deep BCD-Net Using Identical Encoding-Decoding CNN Structures for Iterative Image Recovery
AI总结：
本文提出使用相同编码-解码CNN结构的BCD-Net，通过块坐标下降和交替方向乘子法训练，在低信噪比去噪和极度欠采样MRI中实现更准确的图像恢复。
摘要：
在“极端”计算成像中，收集极度欠采样或噪声污染的测量数据时，在合理的计算时间内获得准确图像具有挑战性。将图像映射卷积神经网络（CNN）融入迭代图像恢复具有解决此问题的巨大潜力。本文1）将编码器和解码器中使用相同卷积核的图像映射CNN融入块坐标下降（BCD）信号恢复方法，2）应用交替方向乘子法训练上述图像映射CNN。我们将所提出的循环网络称为使用相同编码-解码CNN结构的BCD-Net。数值实验表明，对于a）低信噪比图像去噪和b）极度欠采样磁共振成像，与使用不同编码-解码结构的BCD-Net以及使用小波和总变差的传统图像恢复模型相比，所提出的BCD-Net实现了显著更准确的图像恢复。
英文摘要：
In "extreme" computational imaging that collects extremely undersampled or noisy measurements, obtaining an accurate image within a reasonable computing time is challenging. Incorporating image mapping convolutional neural networks (CNN) into iterative image recovery has great potential to resolve this issue. This paper 1) incorporates image mapping CNN using identical convolutional kernels in both encoders and decoders into a block coordinate descent (BCD) signal recovery method and 2) applies alternating direction method of multipliers to train the aforementioned image mapping CNN. We refer to the proposed recurrent network as BCD-Net using identical encoding-decoding CNN structures. Numerical experiments show that, for a) denoising low signal-to-noise-ratio images and b) extremely undersampled magnetic resonance imaging, the proposed BCD-Net achieves significantly more accurate image recovery, compared to BCD-Net using distinct encoding-decoding structures and/or the conventional image recovery model using both wavelets and total variation.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
