Perception for Embodied AI

3 Credits.

This course aims to provide a comprehensive introduction to how robots perceive and understand the physical world through common sensory modalities in embodied systems. Students will study fundamental models of cameras, lidar, event cameras, IMUs, tactile sensors, and proprioception. By the end of the course, students will be able to design and evaluate perception pipelines that connect low-level sensing to high-level robot control and planning. In addition to classical approaches, the course will cover modern advances in robot sensing and perception, including foundation models, learning-based methods, and neural field representations in vision-based robotic systems.