Sleep-like unsupervised replay reduces catastrophic forgetting in artificial neural networks

HIGHLIGHTS

• who: Timothy Tadros from the (UNIVERSITY) have published the Article: Sleep-like unsupervised replay reduces catastrophic forgetting in artificial neural networks, in the Journal: (JOURNAL) of 24/05/2021
• what: For CUB-200, the authors report task 1 performance and task 2 performance separately. performance of sleep phase can be likely improved by incorporating the idea of mini-batches (see Supplementary Information Fig 2 for more detail). While the authors primarily tested a network with only two hidden layers, the analysis of 4-hidden layer networks on MNIST task revealed that SRC can be applied . . .
  

   

  

  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.