Research 2009 - 2011

Simon Liao

Look at Me Now
Teaching Machines to See…Moment by Moment

Simon Liao, Ph.D.
Professor, Applied Computer Science

We are surrounded by artificial intelligence, machines programmed to “think” like us. In reality, these machines are simply processing data, a great deal of complex data, and distinguishing patterns as they emerge from that data. It's a part of machine learning called pattern recognition. It's how the post office scanners read the postal codes on our mail, and it's how that slick new digital camera helps you capture smiles for a perfect picture every time: that little green box that magically appears over faces - and just faces - is pattern recognition.

Our brains distinguish objects regardless of size, position, orientation, or other variables. We know a chair, or even just part of an upside-down chair, when we see one. A computer doesn't have that vision - without the help of Dr. Simon Liao. He uses image analysis through moment measures to identify characteristics of an object that uniquely represent its shape, and writes powerful computer programs to process the data and reproduce the image. He requires significant computer capability to handle the complex calculations.

The author of dozens of theoretical papers, Dr. Liao 's work in pattern recognition provides fundamental techniques that can be used by computer engineers. The results of his research have been applied in health assessment and diagnosis, medical image processing, and robotics.

< Return to Research 2009 - 2011