Skip to content

Dataset · Vihaanakshaay

Satellite benchmark for landslide detection

Vihaanakshaay/LandsatQuake

Geospatial imagery and mask dataset based on 31 landslide inventories from 21 seismic regions for disaster monitoring.

What's inside

The benchmark combines data from 31 landslide inventories across 21 seismic regions worldwide and contains between 1,000 and 10,000 examples. It includes geospatial data, images, and text information for segmentation and object detection tasks. The dataset was presented at the ICLR 2025 workshop on machine learning for Earth observation.

What you can do with it

  • Train landslide segmentation and detection models using satellite imagery
  • Test geospatial neural networks for earthquake impact assessment
Tasks
Object detection, Image segmentation
Size
1K–10K rows
Downloads this month
820
Likes
2

How to load it

from datasets import load_dataset

ds = load_dataset("Vihaanakshaay/LandsatQuake")