MenuBar
DATA SET†
CiFAR-10†
The CIFAR-10 is a labeled dataset of 60,000 images of animals and automobiles collected by Alex Krizhevsky, Vinod Nair, and Geoffery Ninton for training object classification models.
- Featuers:
- 50,000 training images divided into 5 batches of 10,000 images each
- 10,000 testing images of a single batch
- The images are divided into 10 classes; Airplane, Deer, Cat, Dog, Bird, Frog, Horse, ship, Truck.
- Each image is an RGB image of 32x32x3 pixels.
CiFAR-100†
The CIFAR-100 is a labeled dataset of 60,000 images of animals and automobiles collected by Alex Krizhevsky, Vinod Nair, and Geoffery Ninton for training object classification models.
- https://www.cs.toronto.edu/~kriz/cifar.html
- Featuers:
- 50,000 training images divided into 5 batches of 10,000 images each
- 10,000 testing images of a single batch
- The images are divided into 100 classes that are grouped into 20 super classes. Each image is labelled by its class and super class.
- Each image is an RGB image of 32x32x3 pixels.
MNIST†
![[PukiWiki]](image/pukiwiki.png "[PukiWiki]")
