Abstract

A state-of-the-art deep learning framework, HyperReconNet, recover an spectral image from its compressed measurements. However, HyperReconNet does not take the sensing matrix into a account on the training. We propose a residual modbased convolutional neural network to address this limitation.

© 2020 The Author(s)

PDF Article
More Like This
A Deep Learning Approach for Reconstruction in Temporal Compressed Imaging

Linxia Zhang, Jun Ke, and Edmund Y Lam
CW4B.3 Computational Optical Sensing and Imaging (COSI) 2020

Deep Learning for Compressive Spectral Imaging

Xin Miao, Xin Yuan, and Paul Wilford
M3B.3 Digital Holography and Three-Dimensional Imaging (DH) 2019

Model-based Inversion Methods for Compressive Spectral Imaging with Diffractive Lenses

Didem Dogan and Figen S. Oktem
CW4B.4 Computational Optical Sensing and Imaging (COSI) 2020

References

You do not have subscription access to this journal. Citation lists with outbound citation links are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access OSA Member Subscription

Supplementary Material (1)

» Media 1: PDF (14277 KB)