This is an image classification project that has achieved a 93% validation accuracy on Kaggle's Cassava Leaf Disease Classification contest.

The CNN model uses transfer learning and batch normalization to train on an augmented and balanced version of the dataset.

Project by Jenson Choi, Aman Oberoi, and Angeline Xu. UCLA Fall 2021. 🐻