SIEMENS

Visvanathan Ramesh, 48, sits in a dining room that is reminiscent of a rustic German restaurant at a university club in midtown Manhattan. He’s thinking about a place far away: Mumbai (then Bombay), where, in 1984, he was working as a trainee at Larsen&Toubro, an Indian technology company founded by two Danes. “Young employees worked for six weeks each in different departments,” he recalls of the job. “As a consequence one could not get really involved anywhere. So I had plenty of time on my hands — and I read a lot in the company’s research library.”

This not only helped him pass the time but also changed his future — thanks to two articles he read. One was an article in Scientific American magazine about the neurophysiology of vision and how evolving research projects might one day help to develop computer vision. The other article was a story about medical imaging in the journal Proceedings of the IEEE. “Because of those articles, I immediately knew that vision processing was what I wanted to do going forward,” Ramesh says. As a result, he decided to actively pursue his application for a degree in electrical engineering in the United States.

When Ramesh is explaining his career, he often refers to his mother’s determination, courage, and will to work hard. But it was not only this kind of persistence that shaped his path. He also felt a thirst for knowledge — and that brought him to a country that at first was very foreign to him. “There were quite a few cultural differences,” he recalls. “Simple things could become difficult. For example, my visa for the United States said that I had until 1-3-1985 to enter the country — and I figured that meant the first of March. But the American notation puts the month before the day. Because of this, I nearly missed my opportunity to go.”

Ramesh made his dream, which was born in a Mumbai library, a reality. He attended Virginia Tech in southwestern Virginia. He received his PhD from the University of Washington in Seattle, and right afterwards, in 1995, he was offered a job as a research scientist working on automated image processing at Siemens Corporate Research in Princeton, New Jersey. Only a few years later he was promoted and became head of the Real-Time Vision and Modeling department. And his management has led to success. When Ramesh arrived, the department had 30 employees — today there are roughly 150, a large number of them working on medical imaging.

Recognizing Danger. In recent years, Ramesh and his team have created technologies for automatic image processing that are the basis for monitoring systems in subway and train stations, at airports, and in other public spaces. They are also used for cargo and parcel screening. In addition, the majority of driver assistance systems would not be possible without this kind of technology. For example, it can automatically search images for signs of danger — whether a suitcase at an airport gate appears to be unattended or burglars are trying to break into a house.

Ramesh has developed a flexible system architecture that can be adapted to different video analysis applications. What does a video system need in order to monitor a subway system? How do light conditions change? What difficulties are involved in crowd surveillance? “It’s easy to collect vast amounts of image information, but the trick is to integrate this data and interpret it correctly. One important aspect of this is to know what is important and what can be ignored,” he says. Ramesh has been devising statistical models of these influences for a large number of scenarios in recent years. He is responsible for over 120 inventions, 37 of which have been granted patents.

The teams Ramesh works with are located all over the globe — in Bangalore, Munich, Graz, and other locations. “Everyone brings their own perspective to the table,” he says. “whether it’s a methodical German, a Chinese team player or an entrepreneurial American. In such groups, everyone benefits from the strengths of the others — assuming that one is open to these differences. And for that reason, it’s important to have a certain sensitivity for other cultures.”

Ramesh’s global research agenda will certainly continue in the coming years — this time in the interdisciplinary field of cognitive systems. Most recently, Ramesh has been working with a group of German colleagues in the area of artificial intelligence, specifically on discovering how recent discoveries in brain research can be applied to intelligent vision systems — a quarter of a century after he read an article in Scientific American on the same subject.