Near-Duplicate Segments based news web video event mining

Chengde Zhang, Dianting Liu, Xiao Wu, Guiru Zhao, Mei-Ling Shyu, Qiang Peng

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

News web videos uploaded by general users usually include lots of post-processing effects (editing, inserted logo, etc.), which bring noise and affect the similarity comparison for news web video event mining. In this paper, a framework based on the concept of Near-Duplicate Segments (NDSs) which effectively integrates spatial and temporal information is proposed. After each video being divided into segments, those segments from different videos but sharing similar visual content are clustered into groups. Each group is named as an NDS, which infers the latent content relations among videos. The spatial-temporal local features are extracted and used to represent each video segment, which could effectively capture the main content of news web videos and omit the noise such as the disturbance/influence from video editing. Finally, the visual information is integrated with the textual information. The experiment demonstrates that our proposed framework is more effective than several existing methods with a significant improvement.

Original languageEnglish (US)
Pages (from-to)26-35
Number of pages10
JournalSignal Processing
Volume120
DOIs
StatePublished - Mar 1 2016

Fingerprint

Processing
Experiments

Keywords

  • Adaptive Association Rule Mining (AARM)
  • Multiple Correspondence Analysis (MCA)
  • Near-Duplicate Segment (NDS)
  • News web video event mining
  • Spatial-temporal feature

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Control and Systems Engineering
  • Software
  • Signal Processing
  • Computer Vision and Pattern Recognition

Cite this

Near-Duplicate Segments based news web video event mining. / Zhang, Chengde; Liu, Dianting; Wu, Xiao; Zhao, Guiru; Shyu, Mei-Ling; Peng, Qiang.

In: Signal Processing, Vol. 120, 01.03.2016, p. 26-35.

Research output: Contribution to journalArticle

Zhang, Chengde ; Liu, Dianting ; Wu, Xiao ; Zhao, Guiru ; Shyu, Mei-Ling ; Peng, Qiang. / Near-Duplicate Segments based news web video event mining. In: Signal Processing. 2016 ; Vol. 120. pp. 26-35.
@article{bffeacb63eb3473aa7245a28f12bc6cc,
title = "Near-Duplicate Segments based news web video event mining",
abstract = "News web videos uploaded by general users usually include lots of post-processing effects (editing, inserted logo, etc.), which bring noise and affect the similarity comparison for news web video event mining. In this paper, a framework based on the concept of Near-Duplicate Segments (NDSs) which effectively integrates spatial and temporal information is proposed. After each video being divided into segments, those segments from different videos but sharing similar visual content are clustered into groups. Each group is named as an NDS, which infers the latent content relations among videos. The spatial-temporal local features are extracted and used to represent each video segment, which could effectively capture the main content of news web videos and omit the noise such as the disturbance/influence from video editing. Finally, the visual information is integrated with the textual information. The experiment demonstrates that our proposed framework is more effective than several existing methods with a significant improvement.",
keywords = "Adaptive Association Rule Mining (AARM), Multiple Correspondence Analysis (MCA), Near-Duplicate Segment (NDS), News web video event mining, Spatial-temporal feature",
author = "Chengde Zhang and Dianting Liu and Xiao Wu and Guiru Zhao and Mei-Ling Shyu and Qiang Peng",
year = "2016",
month = "3",
day = "1",
doi = "10.1016/j.sigpro.2015.08.002",
language = "English (US)",
volume = "120",
pages = "26--35",
journal = "Signal Processing",
issn = "0165-1684",
publisher = "Elsevier",

}

TY - JOUR

T1 - Near-Duplicate Segments based news web video event mining

AU - Zhang, Chengde

AU - Liu, Dianting

AU - Wu, Xiao

AU - Zhao, Guiru

AU - Shyu, Mei-Ling

AU - Peng, Qiang

PY - 2016/3/1

Y1 - 2016/3/1

N2 - News web videos uploaded by general users usually include lots of post-processing effects (editing, inserted logo, etc.), which bring noise and affect the similarity comparison for news web video event mining. In this paper, a framework based on the concept of Near-Duplicate Segments (NDSs) which effectively integrates spatial and temporal information is proposed. After each video being divided into segments, those segments from different videos but sharing similar visual content are clustered into groups. Each group is named as an NDS, which infers the latent content relations among videos. The spatial-temporal local features are extracted and used to represent each video segment, which could effectively capture the main content of news web videos and omit the noise such as the disturbance/influence from video editing. Finally, the visual information is integrated with the textual information. The experiment demonstrates that our proposed framework is more effective than several existing methods with a significant improvement.

AB - News web videos uploaded by general users usually include lots of post-processing effects (editing, inserted logo, etc.), which bring noise and affect the similarity comparison for news web video event mining. In this paper, a framework based on the concept of Near-Duplicate Segments (NDSs) which effectively integrates spatial and temporal information is proposed. After each video being divided into segments, those segments from different videos but sharing similar visual content are clustered into groups. Each group is named as an NDS, which infers the latent content relations among videos. The spatial-temporal local features are extracted and used to represent each video segment, which could effectively capture the main content of news web videos and omit the noise such as the disturbance/influence from video editing. Finally, the visual information is integrated with the textual information. The experiment demonstrates that our proposed framework is more effective than several existing methods with a significant improvement.

KW - Adaptive Association Rule Mining (AARM)

KW - Multiple Correspondence Analysis (MCA)

KW - Near-Duplicate Segment (NDS)

KW - News web video event mining

KW - Spatial-temporal feature

UR - http://www.scopus.com/inward/record.url?scp=84942278895&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84942278895&partnerID=8YFLogxK

U2 - 10.1016/j.sigpro.2015.08.002

DO - 10.1016/j.sigpro.2015.08.002

M3 - Article

VL - 120

SP - 26

EP - 35

JO - Signal Processing

JF - Signal Processing

SN - 0165-1684

ER -