Dr. Youshan Zhang

By Dave DeFusco

Dr. Youshan Zhang, an expert in artificial intelligence including deep neural networks and computer vision, has joined the Katz School of Science and Health faculty in the master’s program in artificial intelligence.

“I am very excited to become involved in interdisciplinary research and collaborations with researchers from related fields at the Katz School,” said Dr. Zhang, who joined the faculty in August as a tenure-track assistant professor of artificial intelligence and computer science. “I’m interested in investigating new deep networks to solve real problems, especially algorithms that can improve medical imaging to detect diseases.”

Dr. Zhang, who also specializes in machine learning, transfer learning, manifold learning and shape analysis, is currently focusing on improving object detection, classification and labeling using neural networks, which have implications for medical technology and autonomous driving. Neural networks—software inspired by our understanding of how the brain works—can “train” themselves to discover similarities and patterns in data, even when their human creators do not know the patterns exist.

“Youshan is the first full-time computer science and AI faculty hire of the Katz School,” said Dr. Pablo Roldan, director of the program in artificial intelligence and an assistant professor of mathematics. “He is a stellar researcher, and his top-notch teaching skills make him a very valuable addition to the AI program.”

In living organisms, webs of neurons in the brain vastly outperform even the best computer-based networks in perception and pattern recognition, but advances in technology have enabled computers to identify patterns in speech and imagery with great accuracy.

The advantages of this are apparent to consumers, for example, who use Apple’s Siri personal assistant. Current AI programs in new cars can already identify pedestrians and bicyclists from cameras positioned atop the windshield and can stop the car automatically if the driver doesn’t take action to avoid a collision. And, if thousands of cat photos are fed them into a neural network, it can learn to recognize the patterns that define what a cat looks like.

But accurately describing a complex scene requires a deeper representation of what’s going on, capturing how the various objects relate to one another and translating it all into natural-sounding language, which is where Dr. Zhang has honed his expertise.

Over the past seven years, he has authored or co-authored 37 articles on artificial intelligence in scientific journals. As a postdoctoral associate in the Department of Clinical Sciences at Cornell University, he also applied deep learning models in veterinary science to detect the diseases of animals and analyze different animal behaviors.

“Youshan is an exceedingly productive researcher,” said Dr. Paul Russo, dean of the Katz School and vice provost at Yeshiva University. “Students will have the chance to work on cutting-edge science in his lab.”