CNN Architectures(AlexNet,VGG,GoogleNet,ResNet,DenseNet)

AlexNet (2012)

 

The network had a very similar architecture as LeNet by Yann LeCun et al but was deeper, with more filters per layer, and with stacked convolutional layers. It consisted 11x11, 5x5,3x3, convolutions, max pooling, dropout, data augmentation, ReLU activations, SGD with momentum. It attached ReLU activations after every convolutional and fully-connected layer

 

VGGNet (2014)

Similar to AlexNet, only 3x3 convolutions, but lots of filters

1. Convolution using 64 filters
2. Convolution using 64 filters + Max pooling
3. Convolution using 128 filters
4. Convolution using 128 filters + Max pooling
5. Convolution using 256 filters
6. Convolution using 256 filters
7. Convolution using 256 filters + Max pooling
8. Convolution using 512 filters
9. Convolution using 512 filters
10. Convolution using 512 filters + Max pooling
11. Convolution using 512 filters
12. Convolution using 512 filters
13. Convolution using 512 filters + Max pooling
14. Fully connected with 4096 nodes
15. Fully connected with 4096 nodes
16. Output layer with Softmax activation with 1000 nodes

• Convolutions layers (used only 3*3 size )
• Max pooling layers (used only 2*2 size)
• Fully connected layers at end
• Total 16 layers

实际上，一个5x5可以用两个3x3来近似代替，一个7x7可以用三个3x3的卷积核来代替，不仅提升了判别函数的识别能力，而且还减少了参数；如3个3x3的卷积核，通道数为C，则参数为3x(3x3xCxC)=27C2C2，而一个7x7的卷积核，通道数也为C，则参数为(7x7xCxC)=49C2C2。