Cost-Effective Active Learning for Deep Image Classification

• Published in: IEEE transactions on circuits and systems for video technology Vol. 27; no. 12; pp. 2591 - 2600
• Main Authors: Wang, Keze, Zhang, Dongyu, Li, Ya, Zhang, Ruimao, Lin, Liang
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.12.2017; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Active learning; Active learning (AL); Annotations; Artificial neural networks; Classification; Classifiers; deep neural nets; Face recognition; Image classification; incremental learning; Learning; Learning systems; Machine learning; Measurement uncertainty; Neural networks; Object recognition; Representations; Training; Uncertainty; Visualization
• ISSN: 1051-8215; 1558-2205
• DOI: 10.1109/TCSVT.2016.2589879

Recent successes in learning-based image classification, however, heavily rely on the large number of annotated training samples, which may require considerable human effort. In this paper, we propose a novel active learning (AL) framework, which is capable of building a competitive classifier with optimal feature representation via a limited amount of labeled training instances in an incremental learning manner. Our approach advances the existing AL methods in two aspects. First, we incorporate deep convolutional neural networks into AL. Through the properly designed framework, the feature representation and the classifier can be simultaneously updated with progressively annotated informative samples. Second, we present a cost-effective sample selection strategy to improve the classification performance with less manual annotations. Unlike traditional methods focusing on only the uncertain samples of low prediction confidence, we especially discover the large amount of high-confidence samples from the unlabeled set for feature learning. Specifically, these high-confidence samples are automatically selected and iteratively assigned pseudolabels. We thus call our framework cost-effective AL (CEAL) standing for the two advantages. Extensive experiments demonstrate that the proposed CEAL framework can achieve promising results on two challenging image classification data sets, i.e., face recognition on the cross-age celebrity face recognition data set database and object categorization on Caltech-256.