Analytics Vidhya hosted online image classification competition. This notebook shows how to classify images of vehicles based on their images into two categories. Emergency Vehicles like police cars, firetrucks, ambulances and non-emergency vehicles like normal cars, buses and trucks. Dataset consisted of 2352 images, out of which around 1646 were used for training and validation, and 706 were used for testing. The accuracy obtained is pretty satisfying considering the small dataset. More images can be generated using image augmentation techniques to improve accuracy of the overall model.
ankit-21/Emergency-Vehicle-Classification
Classifying a vehicle by seeing their images into an emergency vehicle or not an emergency vehicle
Jupyter Notebook