microsoft/data-centric-satellite-segmentation

Contains implementations of data-centric approaches for improving semantic segmentation on satellite imagery.

31
/ 100
Emerging

This project helps urban planners, environmental analysts, or mapping professionals improve the accuracy of automatically identifying features like buildings or roads in satellite images. It takes raw satellite imagery and associated ground truth maps as input, and outputs refined training data selections and segmentation models that are more precise at classifying objects in the images. This is for professionals who analyze satellite data and need highly accurate land cover maps.

No commits in the last 6 months.

Use this if you need to create more accurate land cover maps or detect specific features from satellite imagery, especially when working with limited labeled data or trying to optimize your training process.

Not ideal if you are looking for an out-of-the-box solution without any technical setup, or if your primary interest is in general object detection rather than precise pixel-level segmentation of satellite images.

satellite-imagery-analysis remote-sensing urban-mapping land-cover-classification geospatial-intelligence
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 8 / 25
Maturity 16 / 25
Community 7 / 25

How are scores calculated?

Stars

56

Forks

3

Language

Python

License

MIT

Last pushed

Apr 10, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/microsoft/data-centric-satellite-segmentation"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.