πŸ‘΅ Classic Networks

Classic Convolutional Neural Networks

Network

First Usage

LeNet-5

Hand written digit classification

AlexNet

ImageNet Dataset

VGG-16

ImageNet Dataset

πŸ”’ LeNet-5

LeNet-5 is a very simple network - By modern standards -. It only has 7 layers;

  • among which there are 3 convolutional layers (C1, C3 and C5)

  • 2 sub-sampling (pooling) layers (S2 and S4)

  • 1 fully connected layer (F6)

  • Output layer

πŸ‘€ Visualization of the network

πŸ™Œ Summary of the network

πŸ›Έ AlexNet

  • Too similar to LeNet-5

  • It has more filters per layer

  • It uses ReLU instead of tanh

  • SGD with momentum

  • Uses dropout instead of regularaization

πŸ‘€ Visualization of the network

πŸ”Ž More Detailed

πŸ™Œ Summary of the network

🌱 VGG-16

πŸ‘€ Visualization of the network

πŸ™Œ Summary of the network

πŸ”Ž More Detailed

😐 Drawbacks

  • It is painfully slow to train (It has 138 million parameters πŸ™„)

πŸ‘©β€πŸ”§ Implementation

🧐 Read More