Research Project: Active Data Collection, Semantic Mapping
Location: 435 Clyde Building
About
Chad Samuelson is a PhD Student in the Electrical and Computer Engineering department with an undergraduate degree in Mechanical Engineering as well as a minor in Computer Science (graduated April 2021). He joined the FRoSt Lab in September 2021.
Chad enjoys working in Machine Learning, Computer Vision, State Estimation, Path Planning, Navigation, and Control Theory. His goal is to become a software engineer for an autonomous vehicle company.
From Spokane, Washington, Chad enjoys anything outdoors especially running, ultimate frisbee, volleyball, pickleball, etc.. Besides sports, Chad enjoys working on personal engineering projects involving 3D printing, mechatronics, automation, and board game design. He also loves spending as much time as possible with his wife.
Research
Chad's research focuses on leveraging semantic knowledge to enable more human-like intelligence in robotic tasks. Chad originally applied a novel visual topic modeling approach with modern Gaussian Processes to help classify underwater imagery data. This classification was performed in a semi-supervised way according to the visual interests of the scientist in a given underwater exploration mission (C. R. Samuelson, 2024).
Chad has since shifted focus to semantic LiDAR mapping in outdoor environments. His research focuses on merging open-set vision language models (VLM), such as CLIP, to the LiDAR domain. He is seeking to build automatic Scene Graphs using this Semantic LiDAR technique allowing for task-driven scene representations. He plans to then expand these representations to a heterogeneous multi-agent scheme.