报告时间：11月11日（星期三） 2:00pm - 3:00pm
报告地点：遥感betway必威西汉姆联官网 附3-202报告厅
主讲人： Alexander C. Loui, Ph.D.
Abstract :
The widespread utilization of digital visual media has motivated many research efforts toward efficient search and retrieval from large media collections. In this seminar, I will present a scene matching algorithm for retrieving complex scenes in consumer image collections. In the absence of reliable GPS information, the location at which a photograph was captured can be described by unique objects in the stationary background that can be matched across images. By applying appropriate constraints to detected keypoint matches between a target and a reference image, complex scenes from a user’s collection can be matched with very good precision. To further improve the computational speed of the algorithm, we propose an efficient approach that incorporates a set of discriminative image features to select representative images for fast location-based image retrieval. These image features are used for classifying images into good vs. bad candidates for scene matching with different image-based metrics. Experimental results show that our approach effectively reduces the computational time by an order of magnitude when compared to the traditional SIFT-, PCA-SIFT-, and SURF-based approaches.
Biography :
Alexander C. Loui obtained his B.A.Sc. (Honors), M.A.Sc, and Ph.D. all in Electrical Engineering from the University of Toronto, Canada. After completing his Ph.D. in 1990, he joined Bellcore in New Jersey as a Member of Technical Staff working on audiovisual compression, multipoint video conferencing, and video-on-demand technologies. From 1996-2013, he was with Kodak Research Labs in Rochester NY as a Principal Scientist in the Image Science Division and later a Principal Investigator at the Computational Science and Technology Lab. He is currently a Senior Principal Scientist and Technical Lead at Kodak Alaris (a spin-off company from Kodak). Dr. Loui has been directing research on multimedia processing and content recognition, video analysis and summarization, digital image management, event detection, image quality assessment, and machine vision applications. He has published over 90 refereed papers in these areas. He is also an Adjunct Professor of the ECE Department at Ryerson University and University of Toronto.
Dr. Loui was Chair (2010-11) of the IEEE Rochester Section, and Chair (2005) of the Rochester Chapter of IEEE Signal Processing Society. Dr. Loui has received a number of awards from Kodak and the R&D community for innovative research on auto-albuming and image management technologies. He is a Kodak Distinguished Inventor (with over 70 granted US patents) and a Fellow of IEEE and SPIE.
