What is it?
|
VisualGREP
research considers the problem of similarity between video sequences. Three
basic questions are raised and (partially) answered. Firstly, at what temporal
duration can video sequences be compared? The frame, shot, scene and video
levels are identified. Secondly, given some image or video feature, what
are the requirements on its distance measure and how can it be "easily"
transformed into the visual similarity desired by the inquirer? Thirdly,
how can video sequences be compared at different levels? A general approach
based on either a set or sequence representa tion with variable degrees
of aggregation is proposed and applied recursively over the different levels
of temporal res olution. It allows the inquirer to fully control the importance
of temporal ordering and duration. The general approach is illustrated
by introducing and discussing some of the many possible image and video
features. Promising experimental results are presented.
|
Publications
|
-
Rainer Lienhart, Wolfgang Effelsberg, and
Ramesh Jain. VisualGREP: A Systematic Method to Compare and Retrieve
Video Sequences. Technical Report TR-97-005, Praktische Informatik
IV, University of Mannheim, October 1997. [Abstract] [PS:
247KB] [PDF]; accepted for publication Kluwer Multimedia Tools and
Applications.
-
Rainer Lienhart, Wolfgang Effelsberg, and
Ramesh Jain. Towards a Visual Grep: A systematic analysis of various
methods to compare video sequences. in Storage and Retrieval for Image
and Video Databases VI, Ishwar. K. Sethi, Ramesh C. Jain, Editors, Proc.
SPIE 3312, pp. 271-282. 1998. [Abstract] [PS
: 60 KB] [PDF
: 140KB]
|
