Giters

huawei-noah / Disout

Code for AAAI 2020 paper, Beyond Dropout: Feature Map Distortion to Regularize Deep Neural Networks (Disout).

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Could you provide more details of "Experiments on Fully Connected Layers"?

jiahy0825 opened this issue · comments

commented

Hi,

Thanks for your great work!

I am reproducing your first experiment recently, but even the baseline method I implemented is very different from the results in the paper. (In your paper, Table 1 says the accuracy of conventional CNN method on CIFAR-10 reaches 81.99%)

My implementation is as consistent as possible with the narrative in your section "Implementation details" of "Experiments on Fully Connected Layers". Maybe I missed some details or tricks (such as padding, momentum, etc.), could you provide more information about this experiment?

commented

The momentum is set to 0.9 following the common implement of SGD.