eduzhai > Applied Sciences > Engineering >

Deep Learning Methods for Solving Linear Inverse Problems Research Directions and Paradigms

  • king
  • (0) Download
  • 20210505
  • Save

... pages left unread,continue reading

Document pages: 60 pages

Abstract: The linear inverse problem is fundamental to the development of variousscientific areas. Innumerable attempts have been carried out to solve differentvariants of the linear inverse problem in different applications. Nowadays, therapid development of deep learning provides a fresh perspective for solving thelinear inverse problem, which has various well-designed network architecturesresults in state-of-the-art performance in many applications. In this paper, wepresent a comprehensive survey of the recent progress in the development ofdeep learning for solving various linear inverse problems. We review how deeplearning methods are used in solving different linear inverse problems, andexplore the structured neural network architectures that incorporate knowledgeused in traditional methods. Furthermore, we identify open challenges andpotential future directions along this research line.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...