/Densenet-on-CIFAR10

CIFAR-10 classification solved with 90% validation accuracy using DenseNet in 50 epochs

Primary LanguageJupyter Notebook

Densenet-on-CIFAR10

This is one of tutorial from my early deep learning days. It is beginner friendly keras tutorial to get you feet wet with basics of deep learning.

It includes concepts of:

  • Image augmentation
  • Transfer learning
  • Learning rate scheduling

Deep learning model used here is DenseNet

The goal is to get best accuracy on validation data (test data) of close to 90% within 50 epochs! Check out the IPython Notebook Densenet_on_CIFAR10.ipynb