Neighborloss: a loss function considering spatial correlation for semantic segmentation of remote sensing image

HIGHLIGHTS

• who: NeighborLoss A Loss Function et al. from the School of Architecture, Chengdu University, Chengdu, China have published the paper: NeighborLoss: A Loss Function Considering Spatial Correlation for Semantic Segmentation of Remote Sensing Image, in the Journal: (JOURNAL)
• what: The main contributions of this paper are summarized. as follows: VOLUME 9, 2021 1) It is proposed that the loss function should also consider the spatial correlation between pixel and its neighborhood pixels, rather than treat each pixel in isolation, and only make the simple idea of plus and average 2) The authors propose a practical . . .
  

   

  

  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.