Improved cross-media relevance model for quick image annotation

doi:10.11772/j.issn.1001-9081.2014.05.1439

Journal of Computer Applications ›› 2014, Vol. 34 ›› Issue (5): 1439-1441.DOI: 10.11772/j.issn.1001-9081.2014.05.1439

Previous Articles Next Articles

Improved cross-media relevance model for quick image annotation

BAO Cuizhu¹,SONG Haiyu¹,NIU Junhai²,XIA Xiu¹,LIN Yaozong¹,WANG Bingfei¹

1. College of Computer Science and Engineering, Dalian Nationalities University, Dalian Liaoning 116600, China;
2. Float Glass Division, Henan Ancai Hi-Tech Company Limited, Anyang Henan 455000, China

Received:2013-11-08 Revised:2013-12-22 Online:2014-05-01 Published:2014-05-30
Contact: SONG Haiyu

快速图像标注的改进跨媒体相关模型

包翠竹¹,宋海玉¹,牛军海²,夏秀¹,林耀宗¹,王炳飞¹

1. 大连民族学院计算机科学与工程学院，辽宁大连 116600
2. 河南安彩高科有限公司浮法玻璃事业部，河南安阳 455000

通讯作者: 宋海玉
作者简介:包翠竹（1990-），女（蒙古族），内蒙古乌兰浩特人，硕士研究生，主要研究方向：图像理解、机器学习；宋海玉（1971-），男，河南安阳人，副教授，博士，主要研究方向：图像理解、计算机视觉、机器学习；牛军海（1974-），男，河南安阳人，工程师，主要研究方向：模式识别、自动控制；夏秀（1990-），女，山东禹城人，硕士研究生，主要研究方向：图像理解；林耀宗（1988-），男，浙江温州人，硕士研究生，主要研究方向：图像理解；王炳飞（1989-），男，河南新乡人，硕士研究生，主要研究方向：图像理解。
基金资助:
国家民委科研资助项目;中央高校基本科研业务费专项

Abstract

Abstract:

To overcome the shortcomings of Cross-Media Relevance Model (CMRM) whose efficiency and effectiveness are low, an improved CMRM was proposed. Based on the improved smoothing method for textual words, the improved CMRM simplified the feature representation and similarity computation which made the measure of relationship between image and image more accurate. The experimental results on the Corel5k dataset show that the proposed approach can significantly improve annotation efficiency. The performance of the improved CMRM is almost three times as good (in terms of mean F1-measure) as original CMRM, also, better than some previously published high quality algorithms such as famous Multiple Bernoulli Relevance Model (MBRM) and Supervised Multiclass Labeling (SML).

摘要：

针对跨媒体相关模型（CMRM）标注效率低、标注效果差的不足，提出了改进的跨媒体相关模型。提出的模型在改进了词汇平滑处理方法的基础之上，通过简洁的图像特征表示方法和相似度计算方法更准确地度量了图像与图像之间的相关性。在Corel5k数据集上的实验结果表明，所提出的改进CMRM标注效率显著提高，性能是原始CMRM的近3倍，而且，也优于高质量的标注模型，如著名的多伯努利相关模型（MBRM）和有指导的多类标签（SML）等模型。

CLC Number:

TP391

BAO Cuizhu SONG Haiyu NIU Junhai XIA Xiu LIN Yaozong WANG Bingfei. Improved cross-media relevance model for quick image annotation[J]. Journal of Computer Applications, 2014, 34(5): 1439-1441.

包翠竹宋海玉牛军海夏秀林耀宗王炳飞. 快速图像标注的改进跨媒体相关模型[J]. 计算机应用, 2014, 34(5): 1439-1441.

References

［1］JEON J, LAVRENKO V, MANMATHA R. Automatic image annotation and retrieval using cross media relevance models ［C］// Proceedings of the 26th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2003:119-126.
［2］LU H, LIU J. Image annotation based on graph learning ［J］. Chinese Journal of Computers, 2008, 31(9) :1629-1539.（卢汉清,刘静.基于图学习的自动图像标注［J］.计算机学报,2008,31(9):1629-1539.）
［3］JELINEK F, MERCER R L. Interpolated estimation of Markov source parameters from sparse data ［C］// Proceedings of the Workshop on Pattern Recognition in Practice. Amsterdam: North Holland, 1980:381-397.
［4］SHI J, MALIK J. Normalized cuts and image segmentation ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000,22(8):888-905.
［5］DENG Y, MANJUNATH B S, SHIN H. Color image segmentation ［C］// Proceedings of the 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE Press, 1999,2:446-451.
［6］SONG H. Research on the automatic image annotation and annotation refinement algorithms ［D］. Changchun: Jilin University, 2012.（宋海玉.自动图像标注及标注改善算法的研究［D］.长春：吉林大学,2012.）
［7］LAVRENKO V, MANMATHA R, JEON J. A model for learning the semantics of pictures ［EB/OL］.［2013-04-21］. http://ciir.cs.umass.edu/pubfiles/mm-46.pdf.
［8］FENG S L, MANMATHA R, LAVRENKO V. Multiple Bernoulli relevance models for image and video annotation ［EB/OL］.［2013-05-05］. http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.58.1879&rep=rep1&type=pdf.
［9］CARNEIRO G, CHAN A B, MORENO P J, et al. Supervised learning of semantic classes for image annotation and retrieval ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007,29(3):394-410.
［10］MAKADIA A, PAVLOVIC V, KUMAR S. Baselines for image annotation ［J］. International Journal of Computer Vision, 2010,90(1):88-105.
［11］MAKADIA A, PAVLOVIC V, KUMAR S. A new baseline for image annotation ［C］// Proceedings of the 10th European Conference on Computer Vision. Berlin: Springer-Verlag, 2008:316-329.

[1]	. Statistically significant sequential patterns mining algorithm under influence [J]. Journal of Computer Applications, 0, (): 0-0.
[2]	. Indoor fall detection algorithm based on Res2Net-YOLACT and feature fusion [J]. Journal of Computer Applications, 0, (): 0-0.
[3]	. One-shot video-based person re-identification with multi-loss learning and joint metric [J]. Journal of Computer Applications, 0, (): 0-0.
[4]	. Network embedding method based on multi-granularity community information [J]. Journal of Computer Applications, 0, (): 0-0.
[5]	. Data stream preference query based on extracted sequence according to temporal condition [J]. Journal of Computer Applications, 0, (): 0-0.
[6]	. Multiscale residual UNet based on attention mechanism to realize breast cancer lesion segmentation [J]. Journal of Computer Applications, 0, (): 0-0.
[7]	. Gene data generation method based on generative adversarial network [J]. Journal of Computer Applications, 0, (): 0-0.
[8]	. Prediction of NOx emission from fluid catalytic cracking unit based on ensemble empirical mode decomposition and long short-term memory network [J]. Journal of Computer Applications, 0, (): 0-0.
[9]	. Fundus vessel segmentation method combined with U-Net and adaptive threshold pulse coupled neural network [J]. Journal of Computer Applications, 0, (): 0-0.
[10]	. EE-GAN：Facial expression recognition method based on generative adversarial network and network integration [J]. Journal of Computer Applications, 0, (): 0-0.
[11]	. Recognition algorithm of vehicle logo on traffic road [J]. Journal of Computer Applications, 0, (): 0-0.
[12]	. Multi-modal deep fusion for false information detection [J]. Journal of Computer Applications, 0, (): 0-0.
[13]	. Point cloud registration algorithm based on residual attention mechanism [J]. Journal of Computer Applications, 0, (): 0-0.
[14]	. Low contrast pulp filament defect detection method of non-woven fabric based on deep feature fusion [J]. Journal of Computer Applications, 0, (): 0-0.
[15]	. Pooling algorithm based on Gaussian function [J]. Journal of Computer Applications, 0, (): 0-0.

Improved cross-media relevance model for quick image annotation

快速图像标注的改进跨媒体相关模型

PDF

Knowledge

Abstract

Cite this article

share this article

References

Related Articles 15

Recommended Articles

Metrics