Joint scene classification and segmentation based on Hidden Markov Model

Jincheng Huang, Zhu Liu, Yao Wang

Research output: Contribution to journalArticle

Abstract

Scene classification and segmentation are fundamental steps for efficient accessing, retrieving and browsing large amount of video data. We have developed a scene classification scheme using a Hidden Markov Model (HMM)-based classifier. By utilizing the temporal behaviors of different scene classes, HMM classifier can effectively classify presegmented clips into one of the predefined scene classes. In this paper, we describe three approaches for joint classification and segmentation based on HMM, which search for the most likely class transition path by using the dynamic programming technique. All these approaches utilize audio and visual information simultaneously. The first two approaches search optimal scene class transition based on the likelihood values computed for short video segment belonging to a particular class but with different search constrains. The third approach searches the optimal path in a super HMM by concatenating HMM's for different scene classes.

Original languageEnglish (US)
Pages (from-to)538-550
Number of pages13
JournalIEEE Transactions on Multimedia
Volume7
Issue number3
DOIs
StatePublished - Jun 2005

Fingerprint

Hidden Markov models
Classifiers
Dynamic programming

Keywords

  • Dynamic programming
  • Hidden Markov model
  • Video analysis
  • Video scene classification
  • Video scene segmentation
  • Video understanding

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Information Systems
  • Computer Graphics and Computer-Aided Design
  • Software

Cite this

Joint scene classification and segmentation based on Hidden Markov Model. / Huang, Jincheng; Liu, Zhu; Wang, Yao.

In: IEEE Transactions on Multimedia, Vol. 7, No. 3, 06.2005, p. 538-550.

Research output: Contribution to journalArticle

@article{cb8ac3cefacd4c8cb23cbde768348c8b,
title = "Joint scene classification and segmentation based on Hidden Markov Model",
abstract = "Scene classification and segmentation are fundamental steps for efficient accessing, retrieving and browsing large amount of video data. We have developed a scene classification scheme using a Hidden Markov Model (HMM)-based classifier. By utilizing the temporal behaviors of different scene classes, HMM classifier can effectively classify presegmented clips into one of the predefined scene classes. In this paper, we describe three approaches for joint classification and segmentation based on HMM, which search for the most likely class transition path by using the dynamic programming technique. All these approaches utilize audio and visual information simultaneously. The first two approaches search optimal scene class transition based on the likelihood values computed for short video segment belonging to a particular class but with different search constrains. The third approach searches the optimal path in a super HMM by concatenating HMM's for different scene classes.",
keywords = "Dynamic programming, Hidden Markov model, Video analysis, Video scene classification, Video scene segmentation, Video understanding",
author = "Jincheng Huang and Zhu Liu and Yao Wang",
year = "2005",
month = "6",
doi = "10.1109/TMM.2005.843346",
language = "English (US)",
volume = "7",
pages = "538--550",
journal = "IEEE Transactions on Multimedia",
issn = "1520-9210",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "3",

}

TY - JOUR

T1 - Joint scene classification and segmentation based on Hidden Markov Model

AU - Huang, Jincheng

AU - Liu, Zhu

AU - Wang, Yao

PY - 2005/6

Y1 - 2005/6

N2 - Scene classification and segmentation are fundamental steps for efficient accessing, retrieving and browsing large amount of video data. We have developed a scene classification scheme using a Hidden Markov Model (HMM)-based classifier. By utilizing the temporal behaviors of different scene classes, HMM classifier can effectively classify presegmented clips into one of the predefined scene classes. In this paper, we describe three approaches for joint classification and segmentation based on HMM, which search for the most likely class transition path by using the dynamic programming technique. All these approaches utilize audio and visual information simultaneously. The first two approaches search optimal scene class transition based on the likelihood values computed for short video segment belonging to a particular class but with different search constrains. The third approach searches the optimal path in a super HMM by concatenating HMM's for different scene classes.

AB - Scene classification and segmentation are fundamental steps for efficient accessing, retrieving and browsing large amount of video data. We have developed a scene classification scheme using a Hidden Markov Model (HMM)-based classifier. By utilizing the temporal behaviors of different scene classes, HMM classifier can effectively classify presegmented clips into one of the predefined scene classes. In this paper, we describe three approaches for joint classification and segmentation based on HMM, which search for the most likely class transition path by using the dynamic programming technique. All these approaches utilize audio and visual information simultaneously. The first two approaches search optimal scene class transition based on the likelihood values computed for short video segment belonging to a particular class but with different search constrains. The third approach searches the optimal path in a super HMM by concatenating HMM's for different scene classes.

KW - Dynamic programming

KW - Hidden Markov model

KW - Video analysis

KW - Video scene classification

KW - Video scene segmentation

KW - Video understanding

UR - http://www.scopus.com/inward/record.url?scp=20444398506&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=20444398506&partnerID=8YFLogxK

U2 - 10.1109/TMM.2005.843346

DO - 10.1109/TMM.2005.843346

M3 - Article

VL - 7

SP - 538

EP - 550

JO - IEEE Transactions on Multimedia

JF - IEEE Transactions on Multimedia

SN - 1520-9210

IS - 3

ER -