CS Department Seminar: John R. Smith, (IBM Research), Semantic Modeling and Retrieval of Image and Video Content

Title: Semantic Modeling and Retrieval of Image and Video Content

Monday, April 21, 2014 ( 2:00 pm to 3:00 pm )

Location: Babbio Center 221

In this talk we describe multiple industry problems requiring content-based analysis and retrieval of images and video across the safety and security, medical, Web, social media and mobile domains.  We present a multi-layer framework for semantic learning and recognition and show how modeling and extraction can be mapped to “big data” distributed computing platforms to enable massive scale image and video analytics.  We describe multiple efforts at IBM, including IBM Multimedia Analysis and Retrieval System (IMARS) and IBM Intelligent Video Analytics (IVA) and demonstrate recent results.  We conclude with future directions for improving multimedia analytics through automated and interactive techniques for learning effective visual features and semantic models.
Dr. John R. Smith is Senior Manager, Intelligent Info. Systems, IBM T. J. Watson Research Center.  He leads IBM's research on multimedia analysis and retrieval including image/video content extraction, semantic modeling, video event detection and social media analysis.  Dr. Smith leads development of IBM Multimedia Analysis and Retrieval System (IMARS) and is PI for IBM's IARPA ALADDIN project.  Dr. Smith conducted some of the earliest work on Web image/video search (WebSEEk and VisualSEEk), which has been highly influential, and has published more than two hundred papers in peer reviewed journals and conferences (16K citations, h-index of 59, i-index of 188).  Dr. Smith is a Fellow of IEEE.


For additional information please contact:
Prof. Gang Hua