Accurate online video tagging via probabilistic hybrid modeling

Jialie Shen, Meng Wang, Tat Seng Chua

Research output: Contribution to journalArticleResearchpeer-review

Abstract

Accurate video tagging has been becoming increasingly crucial for online video management and search. This article documents a novel framework called comprehensive video tagger (CVTagger) to facilitate accurate tag-based video annotation. The system applies both multimodal and temporal properties combined with a novel classification framework with hierarchical structure based on multilayer concept model and regression analysis. The advanced architecture enables effective incorporation of both video concept dependency and temporal dynamics. Using a large-scale test collection containing 50,000 YouTube videos, a set of empirical studies have been carried out and experimental results demonstrate various advantages of CVTagger over the state-of-the-art techniques.

Original languageEnglish
Pages (from-to)99-113
Number of pages15
JournalMultimedia Systems
Volume22
Issue number1
Early online date13 Aug 2014
DOIs
Publication statusPublished - Feb 2016
Externally publishedYes

Fingerprint

Regression analysis
Multilayers

Cite this

Shen, Jialie ; Wang, Meng ; Chua, Tat Seng. / Accurate online video tagging via probabilistic hybrid modeling. In: Multimedia Systems. 2016 ; Vol. 22, No. 1. pp. 99-113.
@article{08d6efcd9729439fa060555c0d70abcb,
title = "Accurate online video tagging via probabilistic hybrid modeling",
abstract = "Accurate video tagging has been becoming increasingly crucial for online video management and search. This article documents a novel framework called comprehensive video tagger (CVTagger) to facilitate accurate tag-based video annotation. The system applies both multimodal and temporal properties combined with a novel classification framework with hierarchical structure based on multilayer concept model and regression analysis. The advanced architecture enables effective incorporation of both video concept dependency and temporal dynamics. Using a large-scale test collection containing 50,000 YouTube videos, a set of empirical studies have been carried out and experimental results demonstrate various advantages of CVTagger over the state-of-the-art techniques.",
keywords = "Online video, Social multimedia, Tagging",
author = "Jialie Shen and Meng Wang and Chua, {Tat Seng}",
year = "2016",
month = "2",
doi = "10.1007/s00530-014-0399-4",
language = "English",
volume = "22",
pages = "99--113",
journal = "Multimedia Systems",
issn = "0942-4962",
publisher = "Springer-Verlag London Ltd.",
number = "1",

}

Accurate online video tagging via probabilistic hybrid modeling. / Shen, Jialie; Wang, Meng; Chua, Tat Seng.

In: Multimedia Systems, Vol. 22, No. 1, 02.2016, p. 99-113.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Accurate online video tagging via probabilistic hybrid modeling

AU - Shen, Jialie

AU - Wang, Meng

AU - Chua, Tat Seng

PY - 2016/2

Y1 - 2016/2

N2 - Accurate video tagging has been becoming increasingly crucial for online video management and search. This article documents a novel framework called comprehensive video tagger (CVTagger) to facilitate accurate tag-based video annotation. The system applies both multimodal and temporal properties combined with a novel classification framework with hierarchical structure based on multilayer concept model and regression analysis. The advanced architecture enables effective incorporation of both video concept dependency and temporal dynamics. Using a large-scale test collection containing 50,000 YouTube videos, a set of empirical studies have been carried out and experimental results demonstrate various advantages of CVTagger over the state-of-the-art techniques.

AB - Accurate video tagging has been becoming increasingly crucial for online video management and search. This article documents a novel framework called comprehensive video tagger (CVTagger) to facilitate accurate tag-based video annotation. The system applies both multimodal and temporal properties combined with a novel classification framework with hierarchical structure based on multilayer concept model and regression analysis. The advanced architecture enables effective incorporation of both video concept dependency and temporal dynamics. Using a large-scale test collection containing 50,000 YouTube videos, a set of empirical studies have been carried out and experimental results demonstrate various advantages of CVTagger over the state-of-the-art techniques.

KW - Online video

KW - Social multimedia

KW - Tagging

UR - http://www.scopus.com/inward/record.url?scp=84955484628&partnerID=8YFLogxK

U2 - 10.1007/s00530-014-0399-4

DO - 10.1007/s00530-014-0399-4

M3 - Article

VL - 22

SP - 99

EP - 113

JO - Multimedia Systems

JF - Multimedia Systems

SN - 0942-4962

IS - 1

ER -