Radiant MLHub

SpaceNet 7

satellite

The SpaceNet 7 Multi-Temporal Urban Development Challenge aims to help address this deficit and develop novel computer vision methods for non-video time series data.

In this challenge, participants will identify and track buildings in satellite imagery time series collected over rapidly urbanizing areas.

The competition centers around a new open source dataset of Planet satellite imagery mosaics, which includes 24 images (one per month) covering ~100 unique geographies.

Dataset ID

spacenet7

Creator

SpaceNet LLC

Contact

https://spacenet.ai/contact-us/

Documentation

Citation

SpaceNet on Amazon Web Services (AWS). “Datasets.” The SpaceNet Catalog. Last modified April 30, 2018. Accessed on [Insert Date]. https://spacenetchallenge.github.io/datasets/datasetHomePage.html.

Python Client example

from radiant_mlhub import Dataset

ds = Dataset.fetch('spacenet7')
for c in ds.collections:
    print(c.id)

Python Client quick-start guide

Source Imagery Collections

Description

SpaceNet 7 Training Source Imagery

License

CC-BY-SA-4.0

Collection ID

sn7_train_source

Download

Description

SpaceNet 7 Test Source Imagery

License

CC-BY-SA-4.0

Collection ID

sn7_test_source

Download

Labels Collection

Description

SpaceNet 7 Training Labels

License

CC-BY-SA-4.0

Collection ID

sn7_train_labels

Download