Max Robotics — Global Service Robot Directory + US Certification

About

Connor Lee is a researcher working at the intersection of computer vision, robotics, and autonomous aerial systems, with a particular focus on thermal imaging and domain adaptation. His most impactful contribution, "Unsupervised RGB-to-Thermal Domain Adaptation via Multi-Domain Attention Network" (2023, 16 citations), introduces a novel framework enabling robots to perform thermal image classification and semantic segmentation without requiring thermal annotations or paired RGB-thermal data — a significant step toward annotation-efficient robot perception. Complementing this work, Lee developed methods for online self-supervised thermal water segmentation and satellite-guided annotation generation for aerial thermal imagery, collectively addressing the persistent challenge of labeled data scarcity in real-world field robotics. His release of the Caltech Aerial RGB-Thermal Dataset (2024, 11 citations) has provided the community with a valuable open benchmark. Earlier work on egocylindrical depth maps for micro air vehicle obstacle avoidance demonstrates his longstanding engagement with aerial robotics. More recently, he has expanded into rehabilitation robotics, investigating hand stiffness and stroke recovery. With a growing citation record and contributions spanning perception, autonomy, and assistive technology, Lee represents a versatile and emerging voice in applied robotics research.