LSUN (Large-scale Scene Understanding Challenge)

The LSUN (Large-scale Scene UNderstanding Challenge) dataset aims to provide a unique benchmark for large-scale scene classification and scene understanding. The LSUN classification dataset contains 10 scene categories such as restaurant, bedroom, kitchen, outdoor church, etc.

The number of images are as follows:

• Training data: 120,000-3 million images per category
• Validation data: contains 300 images
• Test data: Each category contains 1000 images

Example image: