Distinguished Speaker: Andreas Savakis

Photo of Andreas Savakis

Andreas Savakis
Department of Computer Engineering, Rochester Institute of Technology

Intelligent Vision Algorithms for Interactive Display Applications

Andreas Savakis
Department of Computer Engineering, Rochester Institute of Technology

Andreas Savakis spoke on Friday, April 16th, 2010 at 1:00PM in Engineering Building II, Room 1021

Interactive displays offer a wealth of opportunities for applications dealing with human computer interaction, content delivery, entertainment, education, customer service and smart spaces. This talk discusses intelligent vision algorithms developed at RIT's Real Time Vision Lab under the umbrella of interactive display applications. We explore new methodologies for face detection, pose estimation, facial tracking, expression recognition and activity recognition. A variety of approaches are considered ranging from shape models to dimensionality reduction using random projections and manifold learning. The shape context descriptor and active shape models are useful for activity and expression recognition. A random projections transformation is a data-independent, linear dimensionality reduction technique that preserves distances and can be effective for object detection and tracking. Manifold learning is a modeling approach that identifies non-linear structures embedded in the data, and may be employed in numerous applications. In addition to generating elegant signal representations, these dimensionality reduction techniques offer computational efficiency that allows potential extensions on mobile, resource constrained systems.

Dr. Andreas Savakis is Professor and Department Head of Computer Engineering at the Rochester Institute of Technology. He received the B.S. (Summa Cum Laude) and M.S. degrees from Old Dominion University and the Ph.D. from North Carolina State University, all in Electrical Engineering. From 1991-96 he was with the University of Rochester and University of Rochester Medical Center, where he conducted research on pattern recognition, digital radiography, and visual memory. From 1996 to 2000 he was with the Eastman Kodak Company, and worked at the Business Imaging Systems Division, Advanced Development Group and the Kodak Research Labs, Imaging Science Division. Dr. Savakis has been serving as department head of Computer Engineering at RIT since 2000. His current research focuses on novel algorithm development and implementations for object detection and tracking, activity and expression recognition, medical imaging, and processing in multi-camera, resource constrained environments. He has been the primary thesis advisor for 25 students at RIT and his research has generated 11 patents and over 75 publications in journals, conferences and book chapters. Dr. Savakis serves as ABET evaluator for Electrical Engineering and Computer Engineering Programs. His activities were recognized by the IEEE Third Millennium Medal from the IEEE Rochester Section in 2000, and the NYSTAR Technology Transfer Award for Economic Impact in 2006.