Matthew Trang

Papers

Total Citations

H-Index

About

Matthew Trang is a rising star in robotics and artificial intelligence, whose research focuses on bridging the gap between open-set perception and real-time robotic autonomy. His primary contributions lie in developing novel frameworks for 3D scene understanding, enabling robots to build and reason about their environments using task-driven, open-set semantic maps. Trang’s most notable work, *Clio: Real-Time Task-Driven Open-Set 3D Scene Graphs* (2024), has already garnered over 35 citations, reflecting its immediate impact on the field. This pioneering system leverages modern tools like SegmentAnything and CLIP to move beyond traditional closed-set metric-semantic maps, which were limited to a fixed set of semantic classes. Instead, Clio allows robots to dynamically segment and understand their surroundings based on specific tasks, using open-set vocabulary. By integrating real-time performance with class-agnostic segmentation, Trang’s work empowers robots to operate more flexibly in unstructured environments, a critical step toward truly autonomous systems. His research is shaping the next generation of robot perception, making him a key figure to watch in the intersection of computer vision and embodied AI.