Clustering Based Multi-Label Classification for Image Annotation and Retrieval

被引：61

作者：

Nasierding, Gulisong ^{[1
]}

Tsoumakas, Grigorios ^{[3
]}

Kouzani, Abbas Z. ^{[2
]}

机构：

[1] Deakin Univ, Sch Engn, Burwood, Vic 3125, Australia

[2] Deakin Univ, Sch Engn, Geelong, Vic 3217, Australia

[3] Aristotle Univ Thessaloniki, Dept Informat, Thessaloniki GR-54124, Greece

来源：

2009 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS, MAN AND CYBERNETICS (SMC 2009), VOLS 1-9 | 2009年

关键词：

Clustering; multi-label classification; automatic image annotation;

D O I：

10.1109/ICSMC.2009.5346902

中图分类号：

TP3 [计算技术、计算机技术];

学科分类号：

0812 ;

摘要：

This paper presents a novel multi-label classification framework for domains with large numbers of labels. Automatic image annotation is such a domain, as the available semantic concepts are typically hundreds. The proposed framework comprises an initial clustering phase that breaks the original training set into several disjoint clusters of data. It then trains a multi-label classifier from the data of each cluster. Given a new test instance, the framework first finds the nearest cluster and then applies the corresponding model. Empirical results using two clustering algorithms, four multi-label classification algorithms and three image annotation data sets suggest that the proposed approach can improve the performance and reduce the training time of standard multi-label classification algorithms, particularly in the case of large number of labels.

引用

页码：4514 / +

页数：2