As part of the work for his PhD dissertation in electrical engineering, Brian Strope has developed an automatic speech recognition (ASR) system that outperforms state-of-the-art systems in recognizing speech in background noise.

"Aspects of the technology have been around for 20 or 30 years, and for the most part, it works pretty well in really clean situations where you have some good guesses about what the person is going to say and there isn't a lot of interfering sound," Brian says. "When you put it on top of your computer monitor in a real environment with other noises, it doesn't work very well."

Brian's aim--and one goal of the speech signal processing laboratory run by Professor Abeer Alwan--is "to make an ASR work better in real world noisy environments." Brian started by looking at the way people hear. "Humans are really good at being able to recognize speech with competing background sounds," Brian says.

In psychoacoustic experiments, people are placed in soundproof rooms and asked to distinguish speech-like sounds-for example, p, t, k, s, or z--from other noises produced digitally by computer. Brian creates mathematical models to predict how well people will do. Once he gets a model that is close to human performance, he uses it to make better ASR systems.

Brian has done very well in this effort. His work has been published in two journals, and one of several conference papers he has written won the best student paper award at the joint meeting of the Acoustical Societies of America and of Japan in 1996. He also maintained a 3.95 GPA for his graduate coursework.

One of the classes he took was a seminar in comparative physiology, offered by Peter M. Narins, professor of physiological science. There, he used his speech signal processing experience to interpret the physiological experiments in an interesting way. "It is fair to say that the seminar would not have been nearly as successful without Brian's active participation," Dr. Narins said. He "is clearly among the most technically gifted and articulate graduate students I have encountered in my 20 years at UCLA."

Brian's experience in comparative physiology was not the first time that his interests had led him across traditional academic boundaries. As an undergraduate engineering student at Brown University, Brian used his few electives to take music courses. In the computer music program, they started to talk about some signal processing material that was also showing up in his electrical engineering classes. "When those two fields came together," he says, "I knew I had found something I liked."

But right after college, Brian moved in a different direction. At Hewlett-Packard in Fort Collins, Colorado, Brian designed computers for four years before he was ready to move on. "I wanted to work on longer term, more open-ended basic science questions, rather than turn the crank and build the next slightly faster, slightly smaller computer."

His wife, Jennifer Gregory, also wanted to attend graduate school, in clinical psychology. At UCLA, they found an outstanding place to pursue their separate interests. "I've had a lot of freedom to choose what's interesting, to look at what is going to help, and to be able to chase those ideas for a time," Brian says. "I think that's what graduate school is all about." Brian received a UCLA University Fellowship for 1993-1994 and is a member of Tau Beta Pi, the engineering honor society. He expects to complete his PhD this summer.

Over the long-term, he wants to have "more impact on the direction of research, as opposed to turning out the next incremental product." At present, he is looking for faculty positions and opportunities in nonacademic research-oriented agencies that are examining next-generation technical problems. Professors he's had "who come from that environment bring a unique perspective." Brian's long-term goal is to be in a university environment, to "consult part-time and teach most of the time."

"I don't think there's anything I could ever do that would have more impact than to explain something to someone," he says. "I like it when people understand. There's something about that connection that's really satisfying."

Published in Spring 1998 Graduate Quarterly