Multilabel Image Classification Via High-Order Label Correlation Driven Active Learning
- 27 January 2014
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 23 (3), 1430-1441
- https://doi.org/10.1109/tip.2014.2302675
Abstract
Supervised machine learning techniques have been applied to multilabel image classification problems with tremendous success. Despite disparate learning mechanisms, their performances heavily rely on the quality of training images. However, the acquisition of training images requires significant efforts from human annotators. This hinders the applications of supervised learning techniques to large scale problems. In this paper, we propose a high-order label correlation driven active learning (HoAL) approach that allows the iterative learning algorithm itself to select the informative example-label pairs from which it learns so as to learn an accurate classifier with less annotation efforts. Four crucial issues are considered by the proposed HoAL: 1) unlike binary cases, the selection granularity for multilabel active learning need to be fined from example to example-label pair; 2) different labels are seldom independent, and label correlations provide critical information for efficient learning; 3) in addition to pair-wise label correlations, high-order label correlations are also informative for multilabel active learning; and 4) since the number of label combinations increases exponentially with respect to the number of labels, an efficient mining method is required to discover informative label correlations. The proposed approach is tested on public data sets, and the empirical results demonstrate its effectiveness.Keywords
Funding Information
- Australian Government through the Department of Broadband, Communications and the Digital Economy
- Australian Research Council through the ICT Centre of Excellence Program, National ICT Australia
This publication has 31 references indexed in Scilit:
- Crowdsourcing for relevance evaluationACM SIGIR Forum, 2008
- LabelMe: A Database and Web-Based Tool for Image AnnotationInternational Journal of Computer Vision, 2007
- Solving multi-instance problems with classifier ensemble based on constructive clusteringKnowledge and Information Systems, 2006
- Learning multi-label scene classificationPattern Recognition, 2004
- Discriminative Methods for Multi-labeled ClassificationLecture Notes in Computer Science, 2004
- Solving the multiple instance problem with axis-parallel rectanglesArtificial Intelligence, 1997
- Mining association rules between sets of items in large databasesACM SIGMOD Record, 1993
- Queries and concept learningMachine Learning, 1988
- Axiomatic derivation of the principle of maximum entropy and the principle of minimum cross-entropyIEEE Transactions on Information Theory, 1980
- On Information and SufficiencyThe Annals of Mathematical Statistics, 1951