Future Innovators:
NSF CAREER Award Winners

By Geneva Serrer Ringel

One of the highest honors a young faculty member can receive is the National Science Foundation Faculty Early Career Development (CAREER) Award. This prestigious award is given by NSF "to support the early career-development activities of those teacher-scholars who are most likely to become the academic leaders of the 21st century."

Two Ohio State Engineering faculty members have recently been recognized as young professors of this caliber. James W. Davis, assistant professor of computer and information science, and Phil Schniter, assistant professor of electrical engineering, received NSF CAREER awards this year.

"These CAREER award winners are among the best young faculty members in the country," commented Umit Ozkan, associate dean for research in the College of Engineering. "Their excellence strengthens research and innovation within the College of Engineering and provides new opportunities for collaboration."

So, what are the scientific problems these two young promising professors plan to explore?

Thermal cameras images taken outside Ohio State's bookstore.

Computer Vision for Video Surveillance

Imagine being able to determine the activities of individuals with computers and thermal cameras. This is exactly what James Davis, assistant professor of computer and information science, proposes to study with his

CAREER award. He will receive $500,000 from NSF to investigate computer vision methods for video surveillance that can not only detect the presence of people, but also identify their activities.

"People are quite adept at identifying the type of action and amount of effort exerted by a person from subtle visual cues,” explained Davis. “For example, nuances in body motion can reveal how light or heavy a box is for a person to lift or carry. These judgments can even be made from viewing point-light displays or stick figure animations."

Thermal cameras are used for surveillance during night or in low vision circumstances.

James Davis shows Vignesh Kannappan, a computer and information science graduate student, images captured by the infrared camera that will be analyzed by the computer.

"The key to this perception, whether from a machine or human, is the ability to recognize the key visual features that allow us to distinguish human movements and of human movements and their efforts. Current computer vision technology is not yet capable of performing this type of analysis on video surveillance."

Davis plans to have the computer automatically learn the key visual features of different human activities. The curriculum includes collecting an extensive motion capture and video database of several activities at different efforts performed by several people of different age, gender, height, and weight. After identifying the key characteristics of each action, Davis will develop mathematical algorithms that will enable the computer to automatically detect the activities captured in surveillance videos.

So what is the future payoff of this investment in research? Using Davis' algorithms, computers may be able to one day monitor public spaces for suspicious activities. This adds an extra layer of security for our society. The use of thermal cameras also means that night and low vision will no longer provide cover to those wishing to go undetected.

From the cover The thermal images, below, were colored for use on the News in Engineering cover. Picture of model, James Davis, taken with 35 mm camera. Picture of model, James Davis, taken with thermal camera.

Philip Schniter, assistant professor of electrical engineering, will receive $400,000 to investigate new wireless communication schemes and receiver structures that will mitigate fading and interference to ensure a reliable link.

Communication Schemes for Wireless Broadband

Soon, wireless broadband communication will be possible from more than just your home or coffee shop. In the not so distant future, researchers will learn to overcome the problems inherent to mobile communications at high frequency and high data rates. Philip Schniter, assistant professor of electrical engineering, will receive $400,000 to investigate new wireless communication schemes and receiver structures that will mitigate fading and interference to ensure a reliable link.

Wireless signals are highly dynamic and sensitive to their environment. To overcome these sensitivities, information streams are often broken down into components and transmitted via multiple antennas. How the signal is structured, or coded, at the transmitter and then translated at the receiver is controlled by mathematical algorithms.

Currently, algorithms do not yield practical solutions to broadband communication over mobile channels because such signals experience a complicated and quickly changing pattern of signal echoes and interference. So far, only narrowband signals, such as from voice communication, are capable of mobile support. Broadband wireless communication has been gaining momentum, but current solutions require stationary terminals. The simultaneous demand of high frequency, high data rates, and mobile capability creates an extremely challenging problem for wireless broadband communication.

Schniter plans to attack this problem by building on current wireless transmission schemes, such as orthogonal frequency division multiplexing (OFDM). OFDM is used in digital audio and video broadcasting as well as stationary wireless LANs. Extending these ideas, Schniter will develop new algorithms based on non-orthogonal multicarrier signaling schemes, which are better suited to dealing with problems presented by the combination of wide bandwidth and high mobility.

More and more electronic devices have incorporated wireless technology.

Schniter's work may one day pave the way for a new revolution in broadband mobile wireless. In the future, it may be possible to receive and transmit large amounts of data from anywhere, at anytime. This is not limited to the internet. New innovations will result from the ability to transmit and receive complex information through mobile wireless. For instance, it is now possible to feed your pets via wireless broadband signals sent from your PDA to Fido's bowl connected to a home network. In the future, you should be able to confirm that Fido is indeed enjoying his dinner via live digital video.

For more Information:
James Davis, (614) 292-1553, davis.1719@osu.edu
Phil Schniter, (614) 247-6488, Schniter.1@osu.edu