Multi-label image categorization with sparse factor representation

Fuming Sun, Jinhui Tang, Haojie Li, Guo Jun Qi, Thomas S. Huang

Research output: Contribution to journalArticlepeer-review


The goal of multilabel classification is to reveal the underlying label correlations to boost the accuracy of classification tasks. Most of the existing multilabel classifiers attempt to exhaustively explore dependency between correlated labels. It increases the risk of involving unnecessary label dependencies, which are detrimental to classification performance. Actually, not all the label correlations are indispensable to multilabel model. Negligible or fragile label correlations cannot be generalized well to the testing data, especially if there exists label correlation discrepancy between training and testing sets. To minimize such negative effect in the multilabel model, we propose to learn a sparse structure of label dependency. The underlying philosophy is that as long as the multilabel dependency cannot be well explained, the principle of parsimony should be applied to the modeling process of the label correlations. The obtained sparse label dependency structure discards the outlying correlations between labels, which makes the learned model more generalizable to future samples. Experiments on real world data sets show the competitive results compared with existing algorithms.

Original languageEnglish (US)
Article number6705666
Pages (from-to)1028-1037
Number of pages10
JournalIEEE Transactions on Image Processing
Issue number3
StatePublished - Mar 2014
Externally publishedYes


  • Image categorization
  • multilabel
  • sparse

ASJC Scopus subject areas

  • Software
  • Computer Graphics and Computer-Aided Design


Dive into the research topics of 'Multi-label image categorization with sparse factor representation'. Together they form a unique fingerprint.

Cite this