Video Data Modeling Using Sequential Correspondence Hierarchical Dirichlet Processes

Jianfei XUE; Koji EGUCHI

doi:10.1587/transinf.2016MUP0007

Video Data Modeling Using Sequential Correspondence Hierarchical Dirichlet Processes

Jianfei XUE, Koji EGUCHI

Full Text Views

0

Share
Cite this

Summary :

Video data mining based on topic models as an emerging technique recently has become a very popular research topic. In this paper, we present a novel topic model named sequential correspondence hierarchical Dirichlet processes (Seq-cHDP) to learn the hidden structure within video data. The Seq-cHDP model can be deemed as an extended hierarchical Dirichlet processes (HDP) model containing two important features: one is the time-dependency mechanism that connects neighboring video frames on the basis of a time dependent Markovian assumption, and the other is the correspondence mechanism that provides a solution for dealing with the multimodal data such as the mixture of visual words and speech words extracted from video files. A cascaded Gibbs sampling method is applied for implementing the inference task of Seq-cHDP. We present a comprehensive evaluation for Seq-cHDP through experimentation and finally demonstrate that Seq-cHDP outperforms other baseline models.

Publication: IEICE TRANSACTIONS on Information Vol.E100-D No.1 pp.33-41

Publication Date: 2017/01/01

Publicized: 2016/10/07

Online ISSN: 1745-1361

DOI: 10.1587/transinf.2016MUP0007

Type of Manuscript: Special Section PAPER (Special Section on Enriched Multimedia —New Technology Trends in Creation, Utilization and Protection of Multimedia Information—)

Category

Authors

Jianfei XUE
Kobe University
Koji EGUCHI
Kobe University

Keyword

bayesian nonparametric methods, multimedia machine learning, hierarchical Dirichlet processes, topic models

Cite this

Copy

Jianfei XUE, Koji EGUCHI, "Video Data Modeling Using Sequential Correspondence Hierarchical Dirichlet Processes" in IEICE TRANSACTIONS on Information, vol. E100-D, no. 1, pp. 33-41, January 2017, doi: 10.1587/transinf.2016MUP0007.
Abstract: Video data mining based on topic models as an emerging technique recently has become a very popular research topic. In this paper, we present a novel topic model named sequential correspondence hierarchical Dirichlet processes (Seq-cHDP) to learn the hidden structure within video data. The Seq-cHDP model can be deemed as an extended hierarchical Dirichlet processes (HDP) model containing two important features: one is the time-dependency mechanism that connects neighboring video frames on the basis of a time dependent Markovian assumption, and the other is the correspondence mechanism that provides a solution for dealing with the multimodal data such as the mixture of visual words and speech words extracted from video files. A cascaded Gibbs sampling method is applied for implementing the inference task of Seq-cHDP. We present a comprehensive evaluation for Seq-cHDP through experimentation and finally demonstrate that Seq-cHDP outperforms other baseline models.
URL: https://globals.ieice.org/en_transactions/information/10.1587/transinf.2016MUP0007/_p

Copy

@ARTICLE{e100-d_1_33,
author={Jianfei XUE, Koji EGUCHI, },
journal={IEICE TRANSACTIONS on Information},
title={Video Data Modeling Using Sequential Correspondence Hierarchical Dirichlet Processes},
year={2017},
volume={E100-D},
number={1},
pages={33-41},
abstract={Video data mining based on topic models as an emerging technique recently has become a very popular research topic. In this paper, we present a novel topic model named sequential correspondence hierarchical Dirichlet processes (Seq-cHDP) to learn the hidden structure within video data. The Seq-cHDP model can be deemed as an extended hierarchical Dirichlet processes (HDP) model containing two important features: one is the time-dependency mechanism that connects neighboring video frames on the basis of a time dependent Markovian assumption, and the other is the correspondence mechanism that provides a solution for dealing with the multimodal data such as the mixture of visual words and speech words extracted from video files. A cascaded Gibbs sampling method is applied for implementing the inference task of Seq-cHDP. We present a comprehensive evaluation for Seq-cHDP through experimentation and finally demonstrate that Seq-cHDP outperforms other baseline models.},
keywords={},
doi={10.1587/transinf.2016MUP0007},
ISSN={1745-1361},
month={January},}

Copy

TY - JOUR
TI - Video Data Modeling Using Sequential Correspondence Hierarchical Dirichlet Processes
T2 - IEICE TRANSACTIONS on Information
SP - 33
EP - 41
AU - Jianfei XUE
AU - Koji EGUCHI
PY - 2017
DO - 10.1587/transinf.2016MUP0007
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E100-D
IS - 1
JA - IEICE TRANSACTIONS on Information
Y1 - January 2017
AB - Video data mining based on topic models as an emerging technique recently has become a very popular research topic. In this paper, we present a novel topic model named sequential correspondence hierarchical Dirichlet processes (Seq-cHDP) to learn the hidden structure within video data. The Seq-cHDP model can be deemed as an extended hierarchical Dirichlet processes (HDP) model containing two important features: one is the time-dependency mechanism that connects neighboring video frames on the basis of a time dependent Markovian assumption, and the other is the correspondence mechanism that provides a solution for dealing with the multimodal data such as the mixture of visual words and speech words extracted from video files. A cascaded Gibbs sampling method is applied for implementing the inference task of Seq-cHDP. We present a comprehensive evaluation for Seq-cHDP through experimentation and finally demonstrate that Seq-cHDP outperforms other baseline models.
ER -