Hands-On Convolutional Neural Networks with TensorFlow

CIFAR

The CIFAR-10 and CIFAR-100 datasets are small (compared to modern standards) image datasets collected by Alex Krizhevsky, Vinod Nair, and Geoffrey Hinton. These datasets are widely used by the research community for image classification tasks. They are considered challenging, because the image quality is very low and the objects in the images are sometimes partially visible. At the same time, the datasets are convenient due to this small image size, so researchers can quickly produce results on them. CIFAR-100 increases the challenge since there are only a small number of images per class and the number of classes is fairly large. The CIFAR10 and CIFAR100 datasets contain 60,000 images each. Images in both datasets are 32x32x3 RGB color ...