Sherrie Wang, Assistant Professor, Mechanical Engineering & IDSS, MIT


Planetary-Scale Digital Agriculture via Remote Sensing

As the world works toward the UN Sustainable Development Goals, data gaps hinder our ability to measure progress and target interventions. Meanwhile, rapid advances in computer vision and satellite imagery give us the opportunity to extract knowledge about the Earth at planetary scale. Compared to field surveys, satellites offer global coverage at low marginal cost. However, in order to realize the potential of satellite imagery, we must overcome a lack of ground labels, especially in the world’s poorest regions. This talk will cover strategies for mapping agriculture from satellite imagery when ground labels are scarce. In particular, I discuss how transfer learning and weak supervision enable country-wide crop field delineation in India.


Sherrie Wang is an Assistant Professor in a shared position between the Department of Mechanical Engineering and the Institute for Data, Systems, and Society. Her research uses novel data and computational algorithms to monitor our planet and enable sustainable development. Her main application areas are improving agricultural management and mitigating climate change, and she frequently works with satellite imagery, crowdsourced data, and other spatial data. Due to the scarcity of ground truth in many applications and the noisiness of real-world data in general, her methodological work focuses on developing machine learning tools that work within these constraints. Prior to MIT, Sherrie was a Ciriacy-Wantrup Postdoctoral Fellow at UC Berkeley, and she received her PhD in Computational and Mathematical Engineering from Stanford University.