Follow Us:

View Item 
  •   CMRE Open Library Home
  • CMRE Publications
  • Reprints
  • View Item
  •   CMRE Open Library Home
  • CMRE Publications
  • Reprints
  • View Item
JavaScript is disabled for your browser. Some features of this site may not work without it.

Demystifying deep convolutional neural networks for sonar image classification

Thumbnail
Abstract
Deep convolutional neural networks (CNNs) are developed to perform underwater target classification in synthetic aperture sonar (SAS) imagery. The deep networks are trained using a huge database of sonar data collected at sea in various geographical locations. The value of CNN ensemble averaging is highlighted, and the feasibility of sonar sensor transfer learning with CNNs is also demonstrated. An analysis that seeks to demystify deep learning suggests tools for understanding how and why the trained CNNs work so well.

URI
http://hdl.handle.net/20.500.12489/821

Report Number
CMRE-PR-2019-081

Source
In: Proceedings of the 4th Underwater Acoustics Conference, Skiathos, Greece, 3-8 September 2017, pp. 513-520.

Collections
  • Reprints

Date
2019/06

Author(s)
Williams, David P.

Show full item record
CMRE-PR-2019-081.pdf (1.366Mb)

Browse

All of CMRE Open LibraryCommunities & CollectionsBy Issue DateAuthor(s)TitlesSubjectsTypeThis CollectionBy Issue DateAuthor(s)TitlesSubjectsType

My Account

LoginRegister

  • Contact Us
  • Send Feedback
  • Employment
  • Procurement
  • Fact Sheets
  • News Feed
  • Conditions of Use
  • Publications Feed
  • Press Release
  • News Archive
  • STO (Science and Technology Organization)
  • Find us on Facebook
  • Follow us on Twitter
  • Watch us on Youtube
  • Webmail
 

 

© 2018 STO-CMRE
Powered by KnowledgeArc