Dl4ds — deep learning for empirical downscaling

HIGHLIGHTS

• who: METHODS, PAPER and Carlos Alberto, Gomez Gonzalez from the Earth Sciences Department, Barcelona Supercomputing Center, Barcelona, Spain have published the Article: DL4DS u2014 Deep learning for empirical downscaling, in the Journal: (JOURNAL)
• what: As in the case of the Resnet backbone, the authors implement cross-layer skip connections within each dense block and outer skip connections at the backbone section level (Huang et_al, 2017).. In DL4DS, the authors implement a channel attention mechanism based on those of the Squeeze-and-Excitation networks (Hu et_al, 2018) and the Convolutional Block Attention Module (Woo et_al, 2018 . . .
  

   

  

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

  Thanks :)

  Username or Email

  Password

  Remember me

  Lost your password?

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