tetianadadakova / MRI-CNN

Convolutional neural network in TensorFlow for magnetic resonance images reconstruction from frequency domain
72 stars 23 forks source link

convolutional layer #1

Open rockman151 opened 5 years ago

rockman151 commented 5 years ago

Hi,

Your code is pretty good. But I have a question. The paper mentions that the convolutional layers are inspired by Winner-Take-All autoencoders. I tried by myself to add WTA autoencoder to the last convolutional layer but failed. The training cost was very high after applying WTA autoencoder . Do you have any idea about this problem?

tetianadadakova commented 5 years ago

Hi, I haven't tried WTA autoencoders. I am surprised that it affected the cost so much. Have you tried with different hyperparameters?