SpaceNet 6: The Next Big Thing at Topcoder
Topcoder and SpaceNet
Since 2016, Topcoder has been part of SpaceNet, an open innovation project aiming to provide high-quality imagery with state-of-the-art mapping features. The other partners in this innovative project--CosmiQ Works, Digital Globe, and Radiant Solutions --asked the Topcoder community to work on developing better solutions for automated mapping from high-resolution satellite imagery.
Through a series of increasingly complex challenges, Topcoder competitors have provided the SpaceNet team with effective solutions to incorporate into their work.
"The depth, breadth and quality of the responses we've received have been really impactful. Both corporations and research institutions have used the code that we've open-sourced from these challenges... for us that's a huge sign of success."Ryan Lewis, Senior Vice President, CosmiQ Works
What's the impact of these challenges and the resulting solutions? The ability to rapidly create accurate maps is especially important for disaster relief, as maps that reflect changing conditions help first responders work more effectively. Having better mapping technology available in future natural disasters and other crises can save lives and maximize the efficiency of recovery efforts.
Past SpaceNet Challenges: Topcoder's Success
To date, five algorithm challenges have been executed for SpaceNet, with SpaceNet 6 launching soon. Each challenge has drawn between 50 and 100 competitors and resulted in 1,259 total submissions for the SpaceNet series.
The past five SpaceNet challenges increased incrementally in complexity, with competitors trying a multitude of solutions to find winning algorithms. These challenges included:
Building Detection (1 & 2)
Automated Road Network Detection and Route Travel Time Estimation from Satellite Imagery
The work of the Topcoder community has provided many more potential solutions and innovations than SpaceNet's small research team could achieve on their own.
SpaceNet 6: What's New
The SpaceNet 6 challenge will cover Multi-Sensor All-Weather Mapping. The task of SpaceNet 6 is to automatically extract building footprints with computer vision and artificial intelligence (AI) algorithms using a combination of SAR (Synthetic Aperture Radar) and electro-optical imagery datasets. SAR is a unique method of radar that can penetrate cloud cover and provide imagery day or night, and in poor weather conditions (which is particularly important in disaster relief projects). However, the ability to work with SAR imagery is limited by a current lack of relevant open data, which the SpaceNet 6 challenge aims to provide.
The data for the challenge is centered over the port of Rotterdam, the Netherlands. This location, the largest port in Europe, provides a complex dataset for the challenge including around 48,000 buildings, along with vehicles and a variety of boats.
Learn More
Pre-registration for the SpaceNet 6 challenge opens on February 18, 2020. Learn more about SpaceNet 6 here and read the SpaceNet Case Study here.