Prominent AI engineers share their latest work. We will explore the bounds of what is possible with AI and Earth data. Each researcher will share the elements that comprise their modern AI stack -- from managing data, measuring and communicating accuracy, scaling inference, and communicating results.
Speakers: Carolyn Johnston, PhD
Geospatial Data Scientist and Technologist at DevGlobal
Dr. Carolyn Johnston, Ph.D. is a consulting geospatial data scientist and technologist with over 25 years of experience as a scientist, manager, and director of R&D in location technology companies such as Here Technologies, Maxar, and Microsoft. A hands-on technologist with practical knowledge of the open-source geospatial programming stack and deep-learning frameworks, her current interests are in building systems for feature extraction from remote sensing imagery, and in training the next generation of geospatial developers and data scientists. Most recently, she has been the lead data scientist on the RAMP (Replicable AI for Microplanning) project for DevGlobal, where she built an open-source toolkit for rapid derivation of building polygons from satellite imagery over large areas of interest.
Dr. Carolyn Johnston holds a Ph.D. in Mathematics from Louisiana State University, a Master’s degree in Applied Statistics from Colorado State University, and Bachelor’s and Master’s degrees in Mathematics from Binghamton University. She holds patents in image processing, geospatial algorithms, feature extraction from synthetic aperture radar and LiDAR, text and image processing from satellite imagery, and automated mapping for sensor-outfitted autonomous vehicles.
On the agenda
Perspectives from cutting edge applications of EO data in forecasting and prediction.