Abstract | ||
---|---|---|
Image annotation plays an important role in content-based image understanding, various machine learning methods have been proposed to solve this problem. In this paper, label correlation is considered as an undirected bipartite graph, in which each label are correlated by some common hidden topics. As a result, given a label, random walk with restart on the graph supplies a most related label, repeating this precedure leads to a label chain, which keep each adjacent labels pair correlated as maximally as possible. We coordinate the labels chain with its respective classifier training on bottom feature, and guide a classifier chain to annotate an image. The experiment illustrates that our method outperform both the baseline and another popular method. |
Year | DOI | Venue |
---|---|---|
2012 | 10.1007/978-3-642-31537-4_27 | MLDM |
Keywords | Field | DocType |
multi-label image annotation,labels chain,content-based image understanding,label chain,respective classifier training,related label,image annotation,classifier chain,popular method,neighbor pair correlation chain,adjacent label,label correlation | Data mining,Random walk,Computer science,Artificial intelligence,Contextual image classification,Classifier (linguistics),Ensemble learning,Graph,Automatic image annotation,Pattern recognition,Bipartite graph,Correlation,Machine learning | Conference |
Citations | PageRank | References |
1 | 0.36 | 15 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Guang Jiang | 1 | 1 | 0.36 |
Xi Liu | 2 | 36 | 10.08 |
Zhongzhi Shi | 3 | 2440 | 238.03 |