eduzhai > Applied Sciences > Engineering >

Guided Deep Decoder Unsupervised Image Pair Fusion

  • king
  • (0) Download
  • 20210505
  • Save

... pages left unread,continue reading

Document pages: 17 pages

Abstract: The fusion of input and guidance images that have a tradeoff in theirinformation (e.g., hyperspectral and RGB image fusion or pansharpening) can beinterpreted as one general problem. However, previous studies applied atask-specific handcrafted prior and did not address the problems with a unifiedapproach. To address this limitation, in this study, we propose a guided deepdecoder network as a general prior. The proposed network is composed of anencoder-decoder network that exploits multi-scale features of a guidance imageand a deep decoder network that generates an output image. The two networks areconnected by feature refinement units to embed the multi-scale features of theguidance image into the deep decoder network. The proposed network allows thenetwork parameters to be optimized in an unsupervised way without trainingdata. Our results show that the proposed network can achieve state-of-the-artperformance in various image fusion problems.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×