Multimedia Annotation Refinement Based on Contextual Information Diffusion

Authors

Tian Feng, School of Computer and Information Technology, Northeast Petroleum University, Daqing 163318, China;
Fuhua Shang, School of Computer and Information Technology, Northeast Petroleum University, Daqing 163318, China;

Abstract

Abstract: A data driven multimedia annotation refinement method based on dataset contextual information diffusion was proposed. The label contextual graph was constructed, and the label correlation can be diffused on textual label space; Multimedia object content relevant graph was constructed. Label contextual graph and multimedia object content relevant graph were mutually reinforced and formulated into a regularized framework. The proposed method incorporated both multimedia content correlation and label contextual information, and the optimization process was solved by approximate solution algorithm. The experimental results on real world dataset show that the proposed method can obviously improve the annotation performance.

Recommended Citation

Feng, Tian and Shang, Fuhua (2020) "Multimedia Annotation Refinement Based on Contextual Information Diffusion," Journal of System Simulation: Vol. 28: Iss. 11, Article 29.
DOI: 10.16182/j.issn1004731x.joss.201611029
Available at: https://dc-china-simulation.researchcommons.org/journal/vol28/iss11/29

First Page

2860

Revised Date

2016-07-14

DOI Link

https://doi.org/10.16182/j.issn1004731x.joss.201611029

Last Page

2867

CLC

TP391

Recommended Citation

Tian Feng, Shang Fuhua. Multimedia Annotation Refinement Based on Contextual Information Diffusion[J]. Journal of System Simulation, 2016, 28(11): 2860-2867.

DOI

10.16182/j.issn1004731x.joss.201611029