Chih Lai tells of contributions that revolutionize industries.
I was born in Taiwan and received my master’s in computer science and doctorate from Oregon State University. Before I had received my Ph.D., I was recruited by the United Parcel Service Aviation Technology (now Garmin Aviation Technology) to design a GPS-based airplane collision avoidance system. This system received the 1999 Aerospace Industry Pioneer Award as well as six patents; all related to aircraft collision avoidance algorithms. Based on a recent CNN report, the Federal Aviation Administration (FAA) plans to replace the current radar-based system with a similar kind of GPS-based system by 2020.
In the year 2000 I joined the Graduate Programs in Software at the University of St. Thomas. I came to observe that although the surveillance cameras and webcams have become more and more popular, retrieving useful images or videos from those devices is extremely difficult due to the size and unorganized nature of the data. Hence, I decided to apply my previous experience in analyzing large amounts of flight test data to this problem, and expanded the scope of my research from controlling concurrency in real-time systems to recognizing implicit patterns from multimedia data.
In 2001, I began my cooperation with Dr. Dwight Nelson (a former UST biology professor and currently a research manager at Medtronic) in analyzing the circadian rhythms of mice. We analyzed rotation data collected from running wheels in mouse cages. In 2004, we mounted infrared cameras on top of multiple mouse cages in order to continuously monitor every movement of mice over a 10-day period. I then developed new techniques to efficiently process this huge amount of video data (over 460 GB) and automatically categorize similar motions in video into animal activities. These efficient methods are essential in making long video analysis feasible. We have published and presented the results in multiple conferences, proceedings, and journals.
In 2007 I applied those motion analysis techniques to online gaming. In my system, no keyboard, mouse or joystick is required; instead, a player can play online games by holding a flashlight and moving his or her body in front of a webcam. A software program I developed automatically detects, interprets, and converts body motion into inputs required by online games. The idea is to allow players to use body motion to play many free and high quality online games without needing to purchase expensive game consoles and discs.
In 2010 I was invited by a German university to teach and share my research and findings in the area of multimedia information retrieval. I plan on applying my research results to process health-care data, and will offer a Health Care Informatics class in fall 2011 to discuss processing text and images from patient records.
In my research projects, I always engage graduate students. I also have published research results with students. During my tenure at St. Thomas, I have worked with 14 graduate students as well as some undergraduate students. I believe that these research experiences make my graduate students more knowledgeable and better prepared for advanced research projects. Seeing my students succeed in the industry or in a more advanced academic career has been a great privilege of my work so far.