My research focuses on quantitative understanding of visual and cognitive processing in the brain. We address this via modeling and decoding of brain activity evoked under naturalistic conditions.

In our daily life we receive a massive stream of complex and dynamic visual inputs. Our brain processes those inputs to understand the world. This is not a trivial process: the human brain contains several dozens of hierarchically organized cortical areas that process the visual inputs, and those areas cover around one fourth of our entire cortex. Studying how the visual system works gives us a unique opportunity to reveal how the brain analyzes these complex inputs through its functional hierarchy.

We approach this issue via building computational models that can predict brain activity evoked under arbitrary naturalistic conditions. By building such models, we aim to understand the underlying internal representation, cortical mapping, and ultimately the general rules in cortical processing. Currently we use fMRI as our main tool to record human brain activity, but we are also working with data from single-unit recordings.

A quantitative understanding of the brain is crucial in quantitative assessment on how the brain may differ across subjects, different experiences, cognitive conditions, or pathological status. We are to provide a basis for such quantitative assessment for potential future diagnosis. Such models can also be a basis for brain machine interfaces or neuroprosthetics.