In today’s world, robots need a geometrical, semantically rich, and compact 3D representation of their environment. This allows for efficient perception, planning, and navigation, tailored to specific tasks with varying levels of detail, such as navigating to a particular room or locating a specific object. Achieving comprehensive 3D scene understanding requires reasoning about the functional properties and interrelationships of entities within the environment. For a robot to operate fully autonomously, it must be able to function in unfamiliar environments and comprehend scenes even in domains with limited data. This makes it difficult to pre-determine the necessary semantic categories for classification and stresses the importance of open vocabulary solutions.
At the Socially Intelligent Robotics Lab, our research centers on robots operating in complex, human-centered environments, which requires building efficient semantic scene representations that encapsulate scene entities and relations, such as 3D scene graphs and semantic maps. Additionally, we work on dynamic 3D scene reconstruction, semantic SLAM, and efficient 3D geometrical modeling of objects and scenes.
