TY - JOUR
T1 - RDFNet
T2 - Regional Dynamic FISTA-Net for Spectral Snapshot Compressive Imaging
AU - Zhou, Shiyun
AU - Xu, Tingfa
AU - Dong, Shaocong
AU - Li, Jianan
N1 - Publisher Copyright:
© 2015 IEEE.
PY - 2023
Y1 - 2023
N2 - Deep convolutional neural networks have recently shown promising results in compressive spectral reconstruction. Previous methods, however, usually adopt a single mapping function for sparse representation. Considering that different regions have distinct characteristics, it is desirable to apply various mapping functions to adjust different regions' transformations dynamically. With this in mind, we first introduce a regional dynamic way of using Fast Iterative Shrinkage-Thresholding Algorithm (FISTA) to exploit regional characteristics and derive dynamic sparse representations. Then, we propose to unfold the process into a hierarchical dynamic deep network, dubbed RDFNet. The network comprises multiple regional dynamic blocks and corresponding pixel-wise adaptive soft-thresholding modules, respectively in charge of region-based dynamic mapping and pixel-wise soft-thresholding selection. The regional dynamic block guides the network to adjust the transformation domain for different regions. Equipped with the adaptive soft-thresholding, our proposed regional dynamic architecture can also learn appropriate shrinkage scale in a pixel-wise manner. Extensive experiments on both simulated and real data demonstrate that our method outperforms prior state-of-the-arts.
AB - Deep convolutional neural networks have recently shown promising results in compressive spectral reconstruction. Previous methods, however, usually adopt a single mapping function for sparse representation. Considering that different regions have distinct characteristics, it is desirable to apply various mapping functions to adjust different regions' transformations dynamically. With this in mind, we first introduce a regional dynamic way of using Fast Iterative Shrinkage-Thresholding Algorithm (FISTA) to exploit regional characteristics and derive dynamic sparse representations. Then, we propose to unfold the process into a hierarchical dynamic deep network, dubbed RDFNet. The network comprises multiple regional dynamic blocks and corresponding pixel-wise adaptive soft-thresholding modules, respectively in charge of region-based dynamic mapping and pixel-wise soft-thresholding selection. The regional dynamic block guides the network to adjust the transformation domain for different regions. Equipped with the adaptive soft-thresholding, our proposed regional dynamic architecture can also learn appropriate shrinkage scale in a pixel-wise manner. Extensive experiments on both simulated and real data demonstrate that our method outperforms prior state-of-the-arts.
KW - Compressive hyperspectral reconstruction
KW - computational spectral imaging
KW - dynamic neural networks
KW - soft-threshold
UR - http://www.scopus.com/inward/record.url?scp=85147271176&partnerID=8YFLogxK
U2 - 10.1109/TCI.2023.3237175
DO - 10.1109/TCI.2023.3237175
M3 - Article
AN - SCOPUS:85147271176
SN - 2333-9403
VL - 9
SP - 490
EP - 501
JO - IEEE Transactions on Computational Imaging
JF - IEEE Transactions on Computational Imaging
ER -