Robot Perception

3.0 credits · Graduate

An introductory course on robot perception techniques for modeling, fusing, and interpreting heterogeneous and dynamic sensor measurements in the context of robot motion and uncertain environments. The course covers sensor modeling, artificial vision, acoustic sensing, and probabilistic filtering methods. Emphasis is placed on intelligent sensor fusion, object detection and classification, tracking, localization and mapping, exploration, and information-driven motion planning. Algorithms inspired by neural networks, Bayesian networks, graphical models, and information theory are examined. Students investigate perception-driven decision making through benchmark problems such as coverage, target search, tracking, and pursuit-evasion. Applications are drawn from environmental monitoring, surveillance, sensing-and-pursuit games, and human-robot interaction.