Semi-Supervised Image Classification with Self-Paced Cross-Task Networks

Research output: Journal Publications and Reviews (RGC: 21, 22, 62)21_Publication in refereed journalpeer-review

29 Scopus Citations
View graph of relations

Author(s)

  • Si Wu
  • Qiujia Ji
  • Shufeng Wang
  • Zhiwen Yu
  • Yong Xu

Related Research Unit(s)

Detail(s)

Original languageEnglish
Pages (from-to)851-865
Journal / PublicationIEEE Transactions on Multimedia
Volume20
Issue number4
Online published2 Oct 2017
Publication statusPublished - Apr 2018

Abstract

In a semi-supervised setting, direct training of a deep discriminative model on partially labeled images often suffers from overfitting and poor performance, because only a small number of labeled images are available, and errors in label propagation are, in many cases, inevitable. In this paper, we introduce an auxiliary clustering task to explore the structure of the image data, and judiciously weigh unlabeled data to alleviate the influence of ambiguous data on model training. For this purpose, we propose a cross-task network composed of two streams to jointly learn two tasks: classification and clustering. Based on the model predictions, a large number of pairwise constraints can be generated from unlabeled images, and are fed to the clustering stream. Since pairwise constraints encode weak supervision information, the clustering is tolerant of errors in labeling. Unlabeled images are weighted according to the distances to the clusters discovered, and a better discriminative model is trained on the classification stream associated with a weighted softmax loss. Furthermore, a self-paced learning paradigm is adopted to gradually train our deep model from easy examples to difficult ones. Experimental results on widely used image classification datasets confirm the effectiveness and superiority of the proposed approach.

Research Area(s)

  • cross-task network, Image classification, self-paced paradigm, semi-supervised learning

Citation Format(s)

Semi-Supervised Image Classification with Self-Paced Cross-Task Networks. / Wu, Si; Ji, Qiujia; Wang, Shufeng; Wong, Hau-San; Yu, Zhiwen; Xu, Yong.

In: IEEE Transactions on Multimedia, Vol. 20, No. 4, 04.2018, p. 851-865.

Research output: Journal Publications and Reviews (RGC: 21, 22, 62)21_Publication in refereed journalpeer-review