In this project I finetune a CNN model for basic object detection. It also includes special techniques to deal with pre processing and data augmentation of an object detection dataset that consists of bounding boxes in the form of pixel coordinates that is meant to be overlayed on top of images containing objects.