Implementation of DenseNet-121 on the CheXpert Dataset.
This repository is a faithful implementation of the DenseNet-121 image classification model on the chexpert dataset provided by the stanford ML group.
The dataset can be requested through this link : chexpert-stanford ml
A great way to understand how the labelencoder works in assigning the classes could be through this paper Irwin & Rajpurkar et al. at stanford ML.
The repository contains a interactive python notebook which contains the pre-processing, training and testing of the dataset on the chexpert data.
Feel free to create pull requests for any issues or explainations.. would be happy to help