
Stereo Vision (Principles of Robotic Autonomy, Lecture 10)
Marco Pavone, Associate Professor of Aeronautics and Astronautics at Stanford, teaches the tenth lecture of AA274A, Principles of Robotic Autonomy, covering stereo vision as a method for robots to perceive depth. The session builds on the course's progression through perception techniques, explaining how two camera viewpoints can be combined to recover 3D structure from 2D images, a core capability for tasks like obstacle detection and navigation in autonomous systems. Pavone works through the underlying geometry and the computational steps needed to extract depth information, tying the material to the course's companion textbook, Principles of Robot Autonomy. Recorded in Stanford's Autumn 2019 offering of the graduate course, the lecture is aimed at students with a technical background in robotics and assumes familiarity with earlier course material on autonomy fundamentals. It is one installment in a full lecture series available as a playlist, giving viewers a grounded, technical treatment of a key robotic perception problem rather than a general overview.