The Cognitive and Neural System Program investigates human cognitive functions and develop computing systems that are cognitively intelligent. The topics of interests include the understanding of human cognition, adaptive cognitive architecture, cognitive visual understanding, cognitive modeling, and interaction-based learning.

Specifically, we are interested in the following research problems.

1. How can a cognitive architecture integrate a myriad of cognitive functions?
2. How can a cognitive architecture adapt and learn through interaction with the environment?
3. How can cognition be used to guide machine vision?
4. How can a cognitive system support reasoning using sound and robust inference schemes?

As an enlightening example in Cognitive vision, the Kanizsa Square, that merely consists of 4 sets of concentric arcs. Based purely on vision sensors, only the arcs can be detected. However, the cognitive senses resident in humans are able to visualize the phantom square created by the 4 sets of arcs.

Similarly, just as there are cognitive binding circumstances in the visual world, such bindings exist within the worlds of cognitive modeling and interaction-based learning. The binding problems are also evident in web search strategies and surfing patterns. With cognition successfully built into systems, it would be a further breakthrough in the delivery of desired contents over a shorter waiting duration.