Milestone CNN Models Implementations

Implementations of Milestone Convolutional Neural Network Models.

densenet

Motivation

To gain a deeper understanding behind some of the models used in Computer Vision and to practice my Pytorch coding skills, I started implementing popular CNN models.

Responsibilities

  • Re-create CNN models from scratch in Pytorch
  • Attempt to reproduce similar results (accuracies) on the CIFAR10 dataset

Papers implemented so far:

  • VGG (https://arxiv.org/abs/1409.1556)
  • ResNet34 (https://arxiv.org/pdf/1512.03385.pdf)
  • DenseNet (https://arxiv.org/pdf/1608.06993.pdf)

Technologies Used:

  • Languages: Python (Pytorch)