Skip to content

Dataset · timm

RESISC45: remote sensing scene classification benchmark

timm/resisc45

A dataset of 31,500 remote sensing images for 45-class scene classification.

What's inside

Created by Northwestern Polytechnical University (NWPU). Contains 31,500 images across image and text modalities, divided into 45 scene classes with 700 images each. It has no standard default fixed splits.

What you can do with it

  • Training image-classification models
  • Benchmarking computer vision tasks for aerial and satellite data
Tasks
Image classification
Size
10K–100K rows
Downloads this month
2.7K
Likes
8

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("timm/resisc45")