Non-linear dictionary learning with partially labeled data

PATTERN RECOGNITION

卷:48

Article

Shrivastava, Ashish¹  Patel, Vishal M.¹  Chellappa, Rama¹ 

[1] Univ Maryland, Ctr Automat Res, UMIACS, College Pk, MD 20742 USA

关键词: Weakly supervised learning;    Semi-supervised learning;    Kernel methods;    Dictionary learning;    Classification;   

DOI  :  10.1016/j.patcog.2014.07.031

来源: Elsevier

【 摘 要 】

While recent techniques for discriminative dictionary learning have demonstrated tremendous success in image analysis applications, their performance is often limited by the amount of labeled data available for training. Even though labeling images is difficult, it is relatively easy to collect unlabeled images either by querying the web or from public datasets. Using the kernel method, we propose a non-linear discriminative dictionary learning technique which utilizes both labeled and unlabeled data for learning dictionaries in the high-dimensional feature space. Furthermore, we show how this method can be extended for ambiguously labeled classification problem where each training sample has multiple labels and only one of them is correct. Extensive evaluation on existing datasets demonstrates that the proposed method performs significantly better than state of the art dictionary learning approaches when unlabeled images are available for training. (C) 2014 Elsevier Ltd. All rights reserved.

【 授权许可】

Free