eduzhai > Applied Sciences > Engineering >

Advances in Deep Learning for Hyperspectral Image Analysis--Addressing Challenges Arising in Practical Imaging Scenarios

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 25 pages

Abstract: Deep neural networks have proven to be very effective for computer visiontasks, such as image classification, object detection, and semanticsegmentation -- these are primarily applied to color imagery and video. Inrecent years, there has been an emergence of deep learning algorithms beingapplied to hyperspectral and multispectral imagery for remote sensing andbiomedicine tasks. These multi-channel images come with their own unique set ofchallenges that must be addressed for effective image analysis. Challengesinclude limited ground truth (annotation is expensive and extensive labeling isoften not feasible), and high dimensional nature of the data (each pixel isrepresented by hundreds of spectral bands), despite being presented by a largeamount of unlabeled data and the potential to leverage multiple sensors sourcesthat observe the same scene. In this chapter, we will review recent advances inthe community that leverage deep learning for robust hyperspectral imageanalysis despite these unique challenges -- specifically, we will reviewunsupervised, semi-supervised and active learning approaches to image analysis,as well as transfer learning approaches for multi-source (e.g. multi-sensor, ormulti-temporal) image analysis.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×