Open Cities AI Challenge: Results workshop
identify · ai / machine learning · data · emerging technologies
The Open Cities AI Challenge is a global machine learning competition featuring new geospatial data and $15K in prizes to engage the AI expert community to develop better solutions to map buildings in African cities and examine responsible AI use for disaster risk management (DRM). The Challenge dataset represents 10+ diverse cities and regions across Africa with high-resolution drone imagery and building footprint labels created through participatory mapping efforts.
The Challenge competition phase (running from Dec 2019 to Mar 2020 at www.drivendata.org/competitions/60/building-segmentation-disaster-resilience/) will produce high-performing open source machine learning models for building footprint segmentation and thoughtful ideas on responsibly developing and using AI systems for DRM. This workshop session convenes Challenge finalists, organizers, stakeholders, and other ML or disaster risk experts to share results, build mutual learning, and engage the wider Understanding Risk community to explore practical applications of AI solutions for DRM.