Project name: Scene classification
Every day, Google has to filter millions of different landscape photos from around the world, then categorize them into suitable categories for search algorithms. As a result, they needed a computer vision approach to help them categorize landscape photos that internet users had uploaded into the most popular categories.
You are challenged to build a CNN model that classifies landscape images into one of six categories: buildings, forests, rivers, mountains, seas, and streets. (street) with the highest accuracy.
Train, Test and Predict data is separated in each zip file. There are about 14k images in Train, 3k in Test. This data was originally published on datahack for the organization of the Image Classification Contest.