Menu

Understanding the CIFAR-10 Dataset

Understanding the CIFAR-10 Dataset

The CIFAR-10 dataset is one of the most popular datasets used for image classification. It contains 60,000 RGB images divided into 10 different classes. Since the dataset is included in TensorFlow, it can be loaded directly without downloading additional files.

The dataset is divided into:

  • Training Dataset: 50,000 images
  • Testing Dataset: 10,000 images

CIFAR-10 Categories

Label

Category

0Airplane
1Automobile
2Bird
3Cat
4Deer
5Dog
6Frog
7Horse
8Ship
9Truck

Dataset Features

Feature

Description

Total Images60,000
Training Images50,000
Testing Images10,000
Image Size32 × 32 Pixels
Color ChannelsRGB (3 Channels)
Number of Classes10