Targeted data augmentation and hierarchical classification with deep learning for fish species identification in underwater images

HIGHLIGHTS

• who: Abdelouahid Ben Tamou et al. from the CNRS , LabSTICC, France have published the research: Targeted Data Augmentation and Hierarchical Classification with Deep Learning for Fish Species Identification in Underwater Images, in the Journal: (JOURNAL)
• what: The authors propose two Deep Convolutional Neural Network (CNN) approaches for fish species in unconstrained underwater environment. In the first approach the authors use a traditional transfer learning framework and the authors investigate a new technique based on training/validation loss curves for In the second approach the authors propose a CNN to classify fish first into family levels . . .
  

   

  

  If you want to have access to all the content you need to log in!

  Thanks :)

  Username or E-mail:

  If you don't have an account, you can create one here.