Ranking highlight level of movie clips: a template based adaptive kernel SVM method

Zheng Wang, Gaojun Ren, Meijun Sun, Jinchang Ren, Jesse S. Jin

Research output: Contribution to journalSpecial issue

Abstract

This paper looks into a new direction in movie clips analysis – model based ranking of highlight level. A movie clip, containing a short story, is composed of several continuous shots, which is much simpler than the whole movie. As a result, clip based analysis provides a feasible way for movie analysis and interpretation. In this paper, clip-based ranking of highlight level is proposed, where the challenging problem in detecting and recognizing events within clips is not required. Due to the lack of publicly available datasets, we firstly construct a database of movie clips, where each clip is associated with manually derived highlight level as ground truth. From each clip a number of effective visual cues are then extracted. To bridge the gap between low-level features and highlight level semantics, a holistic method of highlight ranking model is introduced. According to the distance between testing clips and selected templates, appropriate kernel function of support vector machine (SVM) is adaptively selected. Promising results are reported in automatic ranking of movie highlight levels.
LanguageEnglish
Pages49-59
Number of pages11
JournalJournal of Visual Languages and Computing
Volume27
Early online date10 Nov 2014
DOIs
Publication statusPublished - 2015

Fingerprint

Support vector machines
Semantics
Testing
Movies
Support Vector Machine
Template
Ranking
Kernel

Keywords

  • video analysis
  • highlight level
  • movie clip
  • template based method
  • adaptive kernel SVM

Cite this

@article{ad1a4d0b0a28409e9813051edb4810b9,
title = "Ranking highlight level of movie clips: a template based adaptive kernel SVM method",
abstract = "This paper looks into a new direction in movie clips analysis – model based ranking of highlight level. A movie clip, containing a short story, is composed of several continuous shots, which is much simpler than the whole movie. As a result, clip based analysis provides a feasible way for movie analysis and interpretation. In this paper, clip-based ranking of highlight level is proposed, where the challenging problem in detecting and recognizing events within clips is not required. Due to the lack of publicly available datasets, we firstly construct a database of movie clips, where each clip is associated with manually derived highlight level as ground truth. From each clip a number of effective visual cues are then extracted. To bridge the gap between low-level features and highlight level semantics, a holistic method of highlight ranking model is introduced. According to the distance between testing clips and selected templates, appropriate kernel function of support vector machine (SVM) is adaptively selected. Promising results are reported in automatic ranking of movie highlight levels.",
keywords = "video analysis, highlight level, movie clip, template based method, adaptive kernel SVM",
author = "Zheng Wang and Gaojun Ren and Meijun Sun and Jinchang Ren and Jin, {Jesse S.}",
year = "2015",
doi = "10.1016/j.jvlc.2014.10.015",
language = "English",
volume = "27",
pages = "49--59",
journal = "Journal of Visual Languages and Computing",
issn = "1045-926X",

}

Ranking highlight level of movie clips : a template based adaptive kernel SVM method. / Wang, Zheng; Ren, Gaojun; Sun, Meijun; Ren, Jinchang; Jin, Jesse S.

In: Journal of Visual Languages and Computing, Vol. 27, 2015, p. 49-59.

Research output: Contribution to journalSpecial issue

TY - JOUR

T1 - Ranking highlight level of movie clips

T2 - Journal of Visual Languages and Computing

AU - Wang, Zheng

AU - Ren, Gaojun

AU - Sun, Meijun

AU - Ren, Jinchang

AU - Jin, Jesse S.

PY - 2015

Y1 - 2015

N2 - This paper looks into a new direction in movie clips analysis – model based ranking of highlight level. A movie clip, containing a short story, is composed of several continuous shots, which is much simpler than the whole movie. As a result, clip based analysis provides a feasible way for movie analysis and interpretation. In this paper, clip-based ranking of highlight level is proposed, where the challenging problem in detecting and recognizing events within clips is not required. Due to the lack of publicly available datasets, we firstly construct a database of movie clips, where each clip is associated with manually derived highlight level as ground truth. From each clip a number of effective visual cues are then extracted. To bridge the gap between low-level features and highlight level semantics, a holistic method of highlight ranking model is introduced. According to the distance between testing clips and selected templates, appropriate kernel function of support vector machine (SVM) is adaptively selected. Promising results are reported in automatic ranking of movie highlight levels.

AB - This paper looks into a new direction in movie clips analysis – model based ranking of highlight level. A movie clip, containing a short story, is composed of several continuous shots, which is much simpler than the whole movie. As a result, clip based analysis provides a feasible way for movie analysis and interpretation. In this paper, clip-based ranking of highlight level is proposed, where the challenging problem in detecting and recognizing events within clips is not required. Due to the lack of publicly available datasets, we firstly construct a database of movie clips, where each clip is associated with manually derived highlight level as ground truth. From each clip a number of effective visual cues are then extracted. To bridge the gap between low-level features and highlight level semantics, a holistic method of highlight ranking model is introduced. According to the distance between testing clips and selected templates, appropriate kernel function of support vector machine (SVM) is adaptively selected. Promising results are reported in automatic ranking of movie highlight levels.

KW - video analysis

KW - highlight level

KW - movie clip

KW - template based method

KW - adaptive kernel SVM

UR - http://www.sciencedirect.com/science/journal/1045926X

U2 - 10.1016/j.jvlc.2014.10.015

DO - 10.1016/j.jvlc.2014.10.015

M3 - Special issue

VL - 27

SP - 49

EP - 59

JO - Journal of Visual Languages and Computing

JF - Journal of Visual Languages and Computing

SN - 1045-926X

ER -