It is a collection of the color image which is commonly used to train machine learning and computer vision algorithms. The CIFAR 10 dataset contains 50000 training images and 10000 validation images such that the images can be classified between 10 different classes.
