Table of Contents
- Introduction to AlexNet
- Walkthrough of AlexNet’s Architecture
- Training of AlexNet
- Summary of AlexNet
1. Introduction to AlexNet
AlexNet was a deep neural network that was developed by Alex Krizhevsky and others in 2012. It was designed to classify images for the ImageNet LSVRC-2010 competition, where it achieved state of the art results [1]. It also worked with multiple GPUs. The AlexNet model contained 11 layers, they were:
- Layer C1: Convolution Layer (96, 11×11)
- Layer S2: Max Pooling Layer (3×3)
- Layer C3: Convolution Layer (256, 5×5)
- Layer S4: Max Pooling Layer (3×3)
- Layer C5: Convolution Layer (384, 3×3)
- Layer C6: Convolution Layer (384, 3×3)
- Layer C7: Convolution Layer (256, 3×3)
- Layer S8: Max Pooling Layer (3×3)
- Layer F9: Fully-Connected Layer (4096)
- Layer F10: Fully-Connected Layer (4096)
- Layer F11: Fully-Connected Layer (1000)
#data-science #artificial-intelligence #machine-learning #convolutional-network