Tuesday, November 5, 2019 - Chicago, Illinois, USA
Title: Creating Global Scale Data Layers from Trillion Pixels using Machine Learning: A Journey of Challenges and Opportunities
Speaker: Dr. Dalton D. Lunga, Oak Ridge National Laboratory
A seismic shift in the commercialization of satellite technology and its rapid deployment is steering the collection of shear volumes of high resolution imagery. As the pervasiveness of machine learning capability in geospatial applications increases, addressing various model generalization challenges is becoming critical. In this talk, I will share lessons learnt at ORNL during our drive to create global scale data layers that are making various societal impact ranging from enabling accurate population distributions, critical infrastructure mapping for damage assessment, and enabling socio-economic stimulation at scale. As a roadmap to future engagements, talk will further present some formidable challenges arising from consideration of trillion pixel capable machine learning systems.
Dr Lunga is currently a lead scientist in machine learning driven geospatial image analytics at ORNL and a member of the ORNL AI Initiative. In this role he deploys machine learning and computer vision techniques in high performance computing environments, focusing on creating imagery-based data layers of interest to various societal problems, e.g., enable accurate population distribution estimates, change-point detection for damage assessment and accurate mapping for socio-economic studies. He currently conducts research and development in machine learning techniques and advanced workflows for handling large volumes of geospatial data. Prior to ORNL, Dr Lunga worked as machine learning research scientist at the council for scientific and industrial research in South Africa on a variety of projects. He received his PhD in electrical and computer engineering from Purdue University, West Lafayette as a Fulbright scholar.