Motion preceptionShape-form-shadingFeature-based attentionPerceptual decision making |
My ReseachIn a broad sense, my research aims to understand how the brain works. But a natural question to ask above all, is this one: is the brain understandable? The answer to this question is going to be a philosophical one or even a religious one. To say either yes or no depends on one's stand point and surely requires more faith than evidence. After all, the number of neurons (the fundamental processing unit) contained in the brain is more than that of the stars in the Galaxy. Those neurons and the connections between them give us senses (smell, taste, touch, see and hear), enable us to think, feel and make sophisticated decisions---the scale of the problem is simply out of measures and out of our imaginations. But, just like physical phenomenon are regular and observing the regularities have led to the discovery of the great laws in physics, human behaviors (governed by various brain functions) contain regularities too. Observing and characterizing them may also lead to deeper understandings of the brain.
Here are a few interesting observations: our five senses (smell, taste, touch, see and hear) give rise to very different sensations. But they all are more sensitive to abrupt changes in stimulation intensities than to constant stimulations. An apple tastes sweeter after you have a bite of a lemon. The New York subway may smell unpleasant to a tourist but not as much to a regular commuter and even less so to the MTA staffs stationed in there. In these observations, different sensory modalities seem to subject to one common computational principle---adaptation, despite of the distinct physical / chemical signals giving rise to various kinds of sensory input. Now consider another observation. In a normal working day morning, you walk out of the house and be well on the way to work. 200 yards later you realize you left your lunch box at home. Will you go back to pick it? If the answer is no, then you are among most of the people that volunteered in this social experiment (privately conducted, data unpublished). Interestingly, similar patterns of behavior also exist on perceptual levels. Inhibition of return, the term that describes the involuntary refusal of the return of one's saccades back to the location visited in the very near past. In this example, there is clearly a common principle that governs processes across different layers of the brain! The list of similar examples will go on... Given that human behaviors contain very rich regularities and are quite predictable, I would argue (boldly and out of faith) that the brain is understandable and studying it will be fruitful. To this end, I have used carefully designed experiments and machine learning techniques to analyze human behavioral data and to generate models to predict certain human behaviors in various perceptual and cognitive tasks. My research has been focused on the area of low level visual perception such as motion perception and more recently has expanded to more higher level cognitions such as attention and decision making. Although the research is limited to only a handful of functions of the brain, the computational principles emerging from these studies appear to apply much more widely. Please click on the links on the left to learn more about each research topic. |