Hidden Markov Models for Video Skim Generation

Authors:
Sergio Benini;Pierangelo Migliorati;Riccardo Leonardi
Affiliations:
University of Brescia, Via Branze 38, 25123, Brescia, Italy;University of Brescia, Via Branze 38, 25123, Brescia, Italy;University of Brescia, Via Branze 38, 25123, Brescia, Italy
Venue:
WIAMIS '07 Proceedings of the Eight International Workshop on Image Analysis for Multimedia Interactive Services
Year:
2007

Citing 0
Cited 5

Multimedia content analysis for consumer electronics

CIVR '08 Proceedings of the 2008 international conference on Content-based image and video retrieval
Feature extraction and clustering for dynamic video summarisation

Neurocomputing
A novel video thumbnail extraction method using spatiotemporal vector quantization

Proceedings of the 3rd international workshop on Automated information extraction in media production
A content-based rapid video playback method using motion-based video time density function and temporal quantization

Proceedings of the 2010 ACM workshop on Social, adaptive and personalized multimedia interaction and access
A smart video player with content-based fast-forward playback

MM '11 Proceedings of the 19th ACM international conference on Multimedia

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper we present a statistical framework based on hidden Markov models (HMMs) for video skimming. A chain of HMMs is used to model subsequent story units: HMM states represent different visual-concepts, transitions model the temporal dependencies in each story unit, and stochastic observations are given by single shots. The skim is generated as an observation sequence, where, in order to privilege more informative segments for entering the skim, dynamic shots are assigned higher probability of observa- tion. The effectiveness of the method is demonstrated on a video set from different kinds of programmes, and results are evaluated in terms of metrics that assess the content rep- resentational value of the obtained video skims.