Constrained utility maximization for generating visual skims

H. Sundaram, Shih Fu Chang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We present a novel algorithm to generate visual skims that do not contain audio, from computable scenes. Visual skims are useful for browsing digital libraries, and for on-demand summaries in set-top boxes. A computable scene is a chunk of data that exhibits consistencies with respect to chromaticity, lighting and sound. First, we define the visual complexity of a shot to be its Kolmogorov complexity. Then, we conduct experiments that help us map the complexity of a shot into the minimum time required for its comprehension. Second, we analyze the grammar of the film language, since it makes the shot sequence meaningful. We achieve a target skim time by minimizing a sequence utility function. It is subject to shot duration constraints, and penalty functions based on sequence rhythm and information loss. This helps us determine individual shot durations as well as the shots to drop. Our user studies show good results on skims with compression rates up to 80%.

Original languageEnglish (US)
Title of host publicationProceedings - IEEE Workshop on Content-Based Access of Image and Video Libraries, CBAIVL 2001
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages124-131
Number of pages8
ISBN (Electronic)0769513549, 9780769513546
DOIs
StatePublished - 2001
Externally publishedYes
EventIEEE Workshop on Content-Based Access of Image and Video Libraries, CBAIVL 2001 - Kauai, United States
Duration: Dec 14 2001 → …

Publication series

NameProceedings - IEEE Workshop on Content-Based Access of Image and Video Libraries, CBAIVL 2001

Other

OtherIEEE Workshop on Content-Based Access of Image and Video Libraries, CBAIVL 2001
Country/TerritoryUnited States
CityKauai
Period12/14/01 → …

ASJC Scopus subject areas

  • General Computer Science

Fingerprint

Dive into the research topics of 'Constrained utility maximization for generating visual skims'. Together they form a unique fingerprint.

Cite this