Home
News
People
Research
Publications
Courses
Seminars
Data & Code
Contact Us
Other Links

Perception: Visual sensor data allows us to measure 3D scene structure, infer the identity and location of objects, and recognize dynamically occurring activities and events.

Action: Active vision systems and smart spaces couple computer vision with robotic actuators to exhibit intiguing, emergent behaviors.

Cognition: Component intelligence systems can learn from training examples and their own exploration. Mathematical representations allow us to model and take advantage of fundamental constraints in the physical world.