Active Imaging for Scarce, Noisy Signals
We build learning and reconstruction methods for radar, sonar, and other active sensing systems that must reason from weak returns, sparse measurements, and imperfect field hardware.
Resilient AI and Grounded Sensing Lab
AI for chaos.
The RAGS Lab develops foundational methods for perception, reasoning, and learning in AI systems that must operate under uncertainty, adapt to changing conditions, and remain reliable in high-stakes real-world environments.
Current research
Current research
Multimodal Inverse Problems
We study how representations transfer across sensing modalities by treating perception as an inverse problem, linking physics, signal structure, and downstream reasoning instead of treating each sensor in isolation.
Reliable Inference Under Pressure
We design AI systems that stay useful when inputs are partial, time is short, and operating conditions shift, with an emphasis on uncertainty, fallback behavior, and deployable compute budgets.
Technology and Policy Co-Design
We shape research questions with operational workflows, policy constraints, and institutional risk in mind so technical prototypes match the environments where they would actually be adopted.
Partner with the lab
Bring hard problems from the field.
We collaborate with first responders, emergency managers, public safety teams, humanitarian operators, students, and researchers who want technology to work in the environments where it will be used.