Research statement (2006)

I am interested in the way we perceive the world. Take an example, we see colors. What exactly is this psychological experience in terms of physical variables and their transformations? How is such a complex and robust computation realized in our brain by organizing a population of neurons that are individually limited in their functionality? Why do we process sensory input in such a specific manner -- does that make sense in light of engineering principles?

I approach these questions by identifying tasks of neural systems, developing models that solve the hypothesized tasks, and testing the models with experimental data. The key to my theoretical approach is to work closely with experimental data by examining published data, collecting data myself, and collaborating with experts of experiments.

The most important feature of this approach is its generality. It leads us to better understand the way one perceives the world with its individual variability caused by different genetic and environmental backgrounds. The same theoretical framework is both informed from and applicable to different sensory modalities, different animals, and artificial systems beyond biology.

Updated research statement (2021)

My overarching research goal is to better understand human experiences and conditions. What exactly are those undeniable subjective experiences of seeing or hearing (e.g., scene analysis) in terms of physical variables and their transformations? How do our brains make sense of raw sensory data that are ambiguous, cluttered, multimodal, and incessantly incoming? How do we interact with an environment that is dynamic, uncertain, and at times so challenging that prior experiences are of little use? How is an immense amount of information about the environment stored in our brains, either innate or learned, and how does it guide our behaviors at need? What causes our minds to go awry, and what interventions can be taken?

To address these questions, I leverage multidisciplinary approaches that integrate machine learning, psychology, biophysics, and neuroscience. My research is characterized by engineering rigor, developing models that not only explain data but also work in practice. And the “theory-experiment closed-loop” is fundamental to my work, allowing theoretical insights to inform experiments and vice versa.

Research topics