Profiles of computer science visiting professors and postdoctoral students
By Melanie Supersano

Editor's note: Computer science is a small department, but it is filled with fresh new energy and talent. Under Yaakov Varol, chair, the department graduated its first doctoral candidate in December (see story page 9) and has attracted visiting professors and postdoctoral fellows from all over the world, five of whom are profiled in this issue of Nevada News.


Chi-Jeng Chang
Visiting professor, Taiwan


Chi-Jeng Chang came to Nevada from the National Taiwan Normal University in September as part of the Nevada-Taiwan exchange program. His background is mainly computer hardware -- digital circuit design, computer architecture. "I have done some processors and CPU circuits design in Field Programmable Gate Array silicon chips, which belongs to the field of very large-scale Integration design," he said.

Chang is also interested in system on chip (SOC) design. "SOC design is actually to put hardware and software together into a single chip with very high gate-count (more than millions of gate-count per chip). So it is better to have some software background for doing SOC design," he explained. Chang is working with Dwight Egbert, professor of computer science, on real-time operating systems.

"I feel lucky to be able to join Dr. [George] Bebis' meetings about research concerning vehicle identification and vehicle tracking problems in the Computer Visions Laboratory," Chang said. "Though computer vision is an area new to me, sometimes it is also hard for me due to much math being involved, but I am very much interested in it because there are many potential SOC applications that could be found in this area."

Jae-Khun Chang
Visiting professor, South Korea


Jae-Khun Chang came to Nevada from South Korea with his wife and two sons last August. They will be here one year.

Chang earned his doctorate in computer science at the University of South Carolina, in 1997, his master's in computer information science at the New Jersey Institute of Technology in 1989, and his bachelor's in architecture from Hanyang University, Seoul, South Korea in 1985.

Chang's main research interest is motion analysis in the computer vision field, which is the one of the strong areas of interest in the Computer Vision Laboratory. "We have a regular meeting every week that is related to a Ford Motor Company project. The goal of the meeting is to find a good algorithm for a real-time vehicle tracking system. This project will be very interesting and helpful for my career," he said.

The Ford-funded project is focusing efforts to develop a robust and reliable real-time, in-vehicle optical detection system that will serve to warn drivers when another vehicle approaches too close. Like an "anti-tailgater" system, or a computerized "backseat driver," the system will alert the driver to dangers on the road. Several factors make on-road vehicle detection very challenging, including variability in scale, location, orientation, and pose.

"I am enjoying Reno life so much," Chang added. "As you know, Reno is the one of the most beautiful places, and my family likes this place."

Ali Erol
Postdoctoral fellow, Turkey


Ali Erol was born in Ankara, Turkey and received his doctoral degree in electrical and electronics engineering from the Middle East Technical University, Ankara, Turkey (METU) in 2001. His dissertation was on fingerprint recognition and his research interests are computer vision -- programming a computer to analyze video, especially of humans, and be able to track and recognize a single individual or an object while in motion -- image processing, pattern recognition and neural networks. While studying for his master's and doctorate degrees, Erol got practical experience as an engineer at a Turkish software company. Before coming to Nevada in August 2002, he worked as an instructor in the Informatics Institute at METU since 1999.

He is working with Associate Professor George Bebis in the Computer Vision Laboratory on a NASA-funded project called "Development of a Nationally Competitive Program in Computer Vision Technologies for Effective Human-Computer Interaction in Virtual Environments."

"My assignment in the project is 3D hand pose estimation and tracking using multiple camera image sequences of the hand," he explained.

Erol enjoys his studies as a post doc. "My studies as a postdoctoral fellow have been very useful and productive. I have enhanced my skills in computer vision and learned about a lot of new topics. Virtual reality is a completely new research topic for me," he said, adding, "The department supplied us a friendly and comfortable working environment."

In addition, Nevada's campus reminds him of home: "Being in Nevada's green campus, just like my old university, METU, is a real pleasure." And Reno's small-town feel is a plus. "Reno is a small, silent and traffic problem-free city. Living here is better than living in a big, crowded city." Yet, "Being close to Lake Tahoe and San Francisco is an important asset," he said.

Haisong Gu
Postdoctoral fellow, PRC


Haisong Gu joined the computer science department at Nevada in January 2002, coming from an industry job as a researcher with Eyematic Interfaces, Inc., a visiting researcher position with the University of Southern California, and a researcher position in the Sensing and Control Laboratory at Matsuhita Electric Works, Ltd., Japan (Panasonic/National).

Gu earned his doctorate in computer vision at Osaka University, Japan.

At Nevada, his research focus is computer vision, working with George Bebis, director of the Computer Vision Laboratory.

Gu recently finished a U. S. Air Force project on fatigue detection. In this project, Gu researched how to detect when a driver or pilot is getting sleepy by detecting whether his or her eyelids close for a certain period of time.

Currently, Gu is partially in charge of a NASA project to explore the advanced technology for gaze estimation and head orientation determination, as well as the Ford Motor Company's vehicle tracking and detection project.

Gu's contributions at Nevada include initiating robust tracking research for facial features, initiating facial expression understanding research, drafting a Department of Defense proposal that got approved, writing three technical papers and partially directing the work of two doctoral candidates and one master's candidate.

His career accomplishments include publishing 25 journal and conference papers; developing state-of-the art techniques for human-computer interaction such as active facial feature tracking, omni-directional vision-based 3D reconstruction of the human body and facial expression recognition; successful commercialization of a stereo vision-based robot and facial tracker and filing four patents.

Gu has been in Reno longer than any city in the United States. "It has very nice living and working conditions," he said. "Computer vision has many potential applications for Nevada industrial interests such as mining and the casino business," he added. "I hope in the near future I can take advantage of my professional and international experiences to make a great contribution to Nevada."

Wenjing Li
Postdoctoral fellow, PRC


Wenjing Li was born in China and earned her doctorate in electronic engineering at the Chinese University of Hong Kong, Hong Kong in 2002. She earned both her bachelor's and master's degrees in computer science and engineering at the Hebei University of Technology, Tianjin, in the People's Republic of China. She came to Nevada last August.

Li is working on an Office of Naval Research project called "Automatic Target Recognition Using Algebraic Functions of Views" with George Bebis, director of the Computer Vision Laboratory.

The main goal of this project is to improve the performance of Automatic Target Recognition (ATR) by developing a more powerful ATR frame work which can handle changes in the appearance of a target more efficiently and robustly. The new framework will be built around a hybrid model of appearance by integrating algebraic functions of views, a powerful mathematical model of geometric appearance, with eigenspace representations, a well known empirical model of appearance which has demonstrated significant capabilities in recognizing complex objects under no occlusion. The project is jointly conducted with the Department of Mechanical Engineering. "I have gained a lot since I came here," Li said. Li hopes that her postdoctoral training at Nevada will help her pursue an academic career.