Multimodal subspace support vector data description

PATTERN RECOGNITION

卷:110

Article

Sohrab, Fahad¹  Raitoharju, Jenni^1,2  Iosifidis, Alexandros³  Gabbouj, Moncef¹ 

[1] Tampere Univ, Fac Informat Technol & Commun Sci, FI-33720 Tampere, Finland

[2] Finnish Environm Inst, Programme Environm Informat, FI-40500 Jyvaskyla, Finland

[3] Aarhus Univ, Dept Engn Elect & Comp Engn, DK-8200 Aarhus, Denmark

关键词: Feature transformation;    Multimodal data;    One-class classification;    Support vector data description;    Subspace learning;   

DOI  :  10.1016/j.patcog.2020.107648

来源: Elsevier

【 摘 要 】

In this paper, we propose a novel method for projecting data from multiple modalities to a new sub-space optimized for one-class classification. The proposed method iteratively transforms the data from the original feature space of each modality to a new common feature space along with finding a joint compact description of data coming from all the modalities. For data in each modality, we define a separate transformation to map the data from the corresponding feature space to the new optimized subspace by exploiting the available information from the class of interest only. We also propose different regularization strategies for the proposed method and provide both linear and non-linear formulations. The proposed Multimodal Subspace Support Vector Data Description outperforms all the competing methods using data from a single modality or fusing data from all modalities in four out of five datasets. (c) 2020 The Authors. Published by Elsevier Ltd.

【 授权许可】

Free