As a principal research scientist at one of the biggest technology companies in the world, Chen Liu has a window into the future. That future, according to the former Beckman Fellow, gets closer every time we perform a search or download a file to our computers.

When the transistor age ushered in electronic signal processing in the 1950s and eventually led to the information technology era of today, it created a vast mountain of data that continues to accumulate in servers and hard drives worldwide. Liu said the next information technology phase is now being built around two key elements: managing all that data and putting it to use, especially in the area of artificial intelligence.

“We are entering an era of how to deal with this data, to use, search for, and retrieve data. To get access to data anywhere, anytime,” Liu said. “The trend will be pattern recognition, machine learning, and data mining.

“Another thing is the intelligence area. Both in academia and industry we are doing things like intelligent cars, intelligent homes. In my lab we develop user interfaces, with a focus on phones, but also on very general research. We definitely need students with new knowledge, like in artificial intelligence and machine learning.”

This, of course, means more opportunities for students in the new data era Liu describes. Those opportunities, he says, won’t just be for computer science or engineering majors.

“When we enter this data era, psychology is definitely a big component of that,” Liu said. “In this lab we have a bunch of scientists in EE and CS, but we find more and more situations where we need to collaborate with linguists and psychologists.”

A student interested in these areas should listen up. Consider that Liu is someone who knows firsthand about being in the right place at the right time and about taking advantage of opportunity. Through his own efforts and a little serendipitous luck, Liu made the most of a one-time chance at the Beckman Institute and eventually earned a coveted position at Motorola Labs.

In 1995 Chen Liu was more than 6,500 miles from Illinois, thinking about how to use the capability certain animals had to pick out individual sounds from a noisy environment in his auditory signal processing research. At the same time a Beckman Institute researcher was pondering how to translate that same ability into something useful for human beings.

It now seems natural for the two quests to become one, but at the time Liu was in Israel doing doctoral work at the Technion-Israel Institute of Technology. He was searching the scientific literature for researchers who could help find a way to deal with the so-called “cocktail party” problem in audio signal processing.

Liu’s aim was to localize and extract a single sound in the presence of multiple interfering sounds – in other words, the kind of environments often found at a cocktail party. That search led him to Beckman’s Al Feng, and to an abrupt change in his life.

“Professor Feng was known worldwide as a specialist in auditory signal processing, especially in directional hearing,” Liu said. “So I contacted him and he said ‘wow we happen to be initiating this kind of project.’ So it was just perfect timing and a perfect match.”

The match turned out to be perfect for both parties. Feng had a postdoctoral research associate to help launch the nascent Intelligent Hearing Aid project, and Liu had a new career path in the United States.

The Beckman Institute got the Intelligent Hearing Aid project started with an initial grant, but there was pressure early on to keep the research and funding on track. Liu said it was a challenge from the very beginning.

“The cocktail party problem was very challenging,” Liu said. “It had been attacked by researchers from around the world for more than four decades. We ventured on a new approach and we had to start from scratch.”

The project started with some equipment and space at Beckman, an algorithm inspired by Feng's seminal research into how frogs delineate sounds in a noisy environment, and Liu. It eventually grew into an interdisciplinary effort involving more than a dozen people and encompassing researchers from electrical and computer engineering, speech and hearing science, and molecular and integrative physiology. The group built an anechoic chamber, testing and recording facilities, and improved and added to the original algorithm.

The Intelligent Hearing Aid project produced high-performance, biologicallyinspired signal processing algorithms that used two directional microphones to extract the desired signal from a noisy environment for use in hearing aids. The rights to the technology were eventually sold to hearing aid manufacturer Phonak.

Feng was the team leader while Liu, who got a Fellows appointment a year into the project, continued to serve as a lead researcher and engineer.

“Chen Liu is an outstanding research scholar, with phenomenal drive, energy and creativity,” Feng said. “His postdoctoral stint at Beckman made it possible for us to formally launch the Intelligent Hearing Aid project. He is a tireless and meticulous researcher, and his work ethic is exemplary. His success in developing the biologicallyinspired computational algorithm paved the way for expansion of the project and subsequent development of other hearing aid and important enabling technologies.”

Liu said the project’s future success was not guaranteed.

“After four months of hard work, during the Christmas of 1995 I finally hit on a new approach and my preliminary tests showed that the idea was doable,” Liu said. “It was based on a hypothetical binaural hearing network structure.

“For the next two more years we kept working, based on the fundamental idea, by improving the algorithms, adding new breakthroughs, and solving implementation issues. Before I left Beckman, we had reached the performance of successfully localizing six simultaneous, equally loud sound sources and extracting only one sound – the capability psychologically observed in the human beings. That was achieved on a computer for the first time.”

Liu joined Motorola in 1998 after a stint as a Beckman Fellow, and after watching the Intelligent Hearing Aid project become a success.

“It was an unforgettable three years for me,” Liu said. “It was one of the most intense growing periods in my research life.”

Liu is now the principal staff research scientist with the Human Interaction Research Lab at Motorola Labs in Schaumburg, Illinois. His group works on aspects of automatic speech recognition, such as robustness, acoustic model training, and multilingual speech recognition. The technology has been incorporated into various Motorola products, including cell phones, home set-top boxes, in-vehicle products, and voice servers.

In addition to having a project go from concept to success on his resume, Liu had the experience of working in an interdisciplinary fashion while at Beckman.

“One thing about the Beckman Institute is the freedom, especially for a Beckman Fellow,” Liu said. “You don’t need to worry about funding; you’re free to do whatever you want. It’s really a rare, precious opportunity for a young scientist.

“Another thing is the interdisciplinary collaborations at the Beckman Institute. A good example is the Intelligent Hearing Aid project. We had faculty members from vastly different fields and we achieved great success. I think it shows the uniqueness of the Beckman Institute.”

In other words, Beckman is the kind of place where young scientists are being prepared for the next technology revolution. The data and know-how are there, Liu says, for this new era to evolve, just as the information technology age has evolved in the last half-century. Now it’s just a matter of making it happen.

“In science fiction 20 or 30 years ago there were things that are now realistic,” Liu said. “All the theoretical and fundamental things have already been laid out and are coming to fruition. There are a lot of things which are technically doable.”

For today’s students interested in being a part of this new era, Liu has some advice. “Broaden your view, try to get various experiences, and think impossible thoughts.”

Those are principles that worked for him.