the goal of the competition is to analyze the data, use data augmentation if needed, and build suitable deep learning models that can recognize 6 categories of natural images (buildings, forest, glacier, mountain, sea and street)