Skip to content

Dataset · gajeshladhar

Earth Observation multimodal satellite data

gajeshladhar/core-five

A large geospatial dataset with temporally and spatially aligned imagery from five satellite sensors.

What's inside

The dataset gathers between 10 and 100 million records of multimodal Earth observation remote sensing data, including Sentinel-2 optical imagery at 10-meter resolution. Data from five different sensors are temporally and spatially aligned into unified spatiotemporal cubes. Accompanying metadata and descriptions are presented in English.

What you can do with it

  • Pretraining multimodal foundational models for satellite imagery analysis
  • Segmentation, object detection, and land cover classification tasks
Tasks
Image segmentation, Image classification, Object detection, Image → text, Image → image, Translation
Languages
English
Size
10M–100M rows
Downloads this month
8.2K
Likes
40

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("gajeshladhar/core-five")