0️⃣ Paper

Visualizing and Understanding Convolutional Networks

1️⃣ Architecture Point

2️⃣ ZFNet Architecture Visualization

3️⃣ Architecture Summary

Layer # Filters / neurons Filter Size Stride Padding Size of Feature Map
Input - - - - 224 x 224 x 3
Convolution 1 96 7 x 7 2 1 110 x 110 x 96
Max Pool 1 - 3 x 3 2 1 55 x 55 x 96
Convolution 2 256 5 x 5 2 - 26 x 26 x 256
Max Pool 2 - 3 x 3 2 1 13 x 13 x 256
Convolution 3 384 3 x 3 1 1 13 x 13 x 384
Convolution 4 384 3 x 3 1 1 13 x 13 x 384
Convolution 5 256 3 x 3 1 1 13 x 13 x 256
Max Pool 3 - 3 x 3 2 - 6 x 6 x 256
Dense(FC) 1 4096 1 x 4096
Dense(FC) 2 4096 1 x 4096
Softmax $C$ (General. 1000) 1 x 1000

4️⃣ Implement Code