CCNY PRISM Lecture Series on

Computer Vision, Robotics and Human-Computer Interaction


Title: Video Mosaicing for Non-Chronological Time Editing

Professor Shmuel Peleg
School of Computer Science and Engineering
The Hebrew University of Jerusalem

Date: Wednesday, August 2, 2006
Time: 2:00 pm - 3:00 pm
Location: Steinman Hall, Room T-623 (EE Conference Room)

ABSTRACT

Time does not have to remain chronological in the new approach to video editing, and frames of the new video are composed of frame parts captured at different times. The elimination of the "chronological time" constraint enables the creation of new and useful effects: dynamic panoramic images, accelerating competitors in sport events, panoramic stereo images, making a long video shorter, and more. Many video clips will demonstrate the possible effects.

All effects share a common methodology: the frames of a video clip are represented as a 3D space-time volume. A new video is generated by sweeping a "time front surface" through this space-time volume. Each "time front" corresponds to a new output frame generated from the input video, and the evolution of the time front creates the output video.

Biography

Shmuel Peleg received the BSc degree in mathematics from The Hebrew University of Jerusalem, Israel, in 1976 and the MSc and PhD degrees in computer science from the University of Maryland, College Park, in 1978 and 1979, respectively. He has been a faculty member at the Hebrew University of Jerusalem since 1980, and was the chairman of the school of computer science from 1989 to 1991.
He has held visiting positions at the University of Maryland, New York University, and the Sarnoff Corporation.

Shmuel published over 130 technical papers in computer vision. His research covered pyramid representation, image enhancement, motion analysis, and panoramic mosaicing. He was issued several patents which provided the technical foundations to three startup companies: VideoBrush (USA), Emaki (Japan), and HumanEyes (Israel).

The lecture series is supported by the CCNY Grove School of Engineering, and by the CCNY Center of the Perceptual Robotics, Intelligent Sensors and Machines (PRISM) through a grant from the NSF CISE Computing Research Infrastructure (CRI)  program.