Mediaspace scheduled maintenance: Aug 25, 2026 07:00 - 12:00 AM. During this time, videos will be temporarily unavailable. Check status updates.
This lecture discusses the role of feedback in visual intelligence, emphasizing its importance in understanding perception and adaptation. The instructor begins by reviewing foundational concepts such as bottom-up and top-down processing, highlighting how feedback mechanisms influence visual perception. Various examples, including the Necker Cube and the Duck-Rabbit illusion, illustrate how expectations shape our interpretation of ambiguous stimuli. The lecture then transitions to computational feedback methods, detailing inner and outer loop feedback systems. The inner loop focuses on processing without external input, while the outer loop incorporates feedback from the environment, akin to control systems like PID controllers. The instructor presents recent advancements in rapid motor adaptation for robots, showcasing how feedback allows for real-time adjustments in diverse environments. The discussion culminates in the application of feedback in neural networks, particularly in adapting to distribution shifts in data. Overall, the lecture provides a comprehensive overview of how feedback mechanisms enhance both human and machine perception and action.
This video is available exclusively on Mediaspace for a restricted audience. Please log in to MediaSpace to access it if you have the necessary permissions.
Watch on Mediaspace