Stanford Seminar - Multi-Sensory Neural Objects: Modeling, Inference, and Applications in Robotics

Explore the cutting-edge developments in multi-sensory neural object representations and their applications in robotics through this Stanford seminar. Delve into the research of Jiajun Wu, Assistant Professor of Computer Science at Stanford University, as he presents recent efforts in building object-centric and multi-sensory neural representations. Learn about technical innovations, advances in scaling up these representations, methods for inferring category-agnostic neural object representations, and systems adopting these representations for robotic manipulation. Gain insights into the potential of these neural representations for improving graphics, vision tasks, and robotic applications. Discover how this research draws inspiration from human cognition to enhance machine perception, reasoning, and interaction with the physical world.