LeNet

LeNet-5 ,以用于实现手写识别的7CNN(不包含输入层)为例,以下为LeNet-5的示意图:

   
   
   

LeNet_第1张图片

Input Layer1*32*32图像

Conv1 Layer:包含6个卷积核,kernal size5*5parameters:5*5+1*6=156

Subsampling Layeraverage poolingsize2*2

                                Activation Functionsigmoid

Conv3 Layer:包含16个卷积核,kernal size5*5 ->16Feature Map

Subsampling Layeraverage poolingsize2*2

Conv5 Layer:包含120个卷积核,kernal size5*5

Fully Connected LayerActivation Functionsigmoid

Output LayerGaussian connections

你可能感兴趣的:(tensorflow)