About the project

The main goal of this project is to discover and explain how internals of convolution layers and fully connected layers provide recognition of depicted information. Original net design of the AlexNet has been described in the NIPS 2012 publication. This project examines contemporary version of implementation contained in PyTorch™ example series.

Library

• 📜 ImageNet Classification with Deep Convolutional Neural Networks, NIPS 2012
• 📚 AlexNet, Wikipedia
• 📑 PyTorch documentation

ImageNet™ datasets

• 📊 Large Scale Visual Recognition Challenge 2012 (ILSVRC2012)
• 📑 Synsets retained from ILSVRC2011
• 💿 wget https://image-net.org/data/ILSVRC/2012/ILSVRC2012_img_train.tar # dataset for training, 137 GB
• 💿 wget https://image-net.org/data/ILSVRC/2012/ILSVRC2012_img_val.tar # dataset for testing, 6 GB