TY - JOUR
T1 - Unsupervised Multi-view Feature Extraction with Dynamic Graph Learning
AU - Shi, Dan
AU - Zhu, Lei
AU - Cheng, Zhiyong
AU - Li, Zhihui
AU - Zhang, Huaxiang
PY - 2018/10
Y1 - 2018/10
N2 - Graph-based multi-view feature extraction has attracted much attention in literature. However, conventional solutions generally rely on a manually defined affinity graph matrix, which is hard to capture the intrinsic sample relations in multiple views. In addition, the graph construction and feature extraction are separated into two independent processes which may result in sub-optimal results. Furthermore, the raw data may contain adverse noises that reduces the reliability of the affinity matrix. In this paper, we propose a novel Unsupervised Multi-view Feature Extraction with Dynamic Graph Learning (UMFE-DGL) to solve these limitations. We devise a unified learning framework which simultaneously performs dynamic graph learning and the feature extraction. Dynamic graph learning adaptively captures the intrinsic multiple view-specific relations of samples. Feature extraction learns the projection matrix that could accordingly preserve the dynamically adjusted sample relations modelled by graph into the low-dimensional features. Experimental results on several public datasets demonstrate the superior performance of the proposed approach, compared with state-of-the-art techniques.
AB - Graph-based multi-view feature extraction has attracted much attention in literature. However, conventional solutions generally rely on a manually defined affinity graph matrix, which is hard to capture the intrinsic sample relations in multiple views. In addition, the graph construction and feature extraction are separated into two independent processes which may result in sub-optimal results. Furthermore, the raw data may contain adverse noises that reduces the reliability of the affinity matrix. In this paper, we propose a novel Unsupervised Multi-view Feature Extraction with Dynamic Graph Learning (UMFE-DGL) to solve these limitations. We devise a unified learning framework which simultaneously performs dynamic graph learning and the feature extraction. Dynamic graph learning adaptively captures the intrinsic multiple view-specific relations of samples. Feature extraction learns the projection matrix that could accordingly preserve the dynamically adjusted sample relations modelled by graph into the low-dimensional features. Experimental results on several public datasets demonstrate the superior performance of the proposed approach, compared with state-of-the-art techniques.
KW - Dynamic graph learning
KW - Intrinsic sample relations
KW - Multi-view feature extraction
UR - http://www.scopus.com/inward/record.url?scp=85054274833&partnerID=8YFLogxK
U2 - 10.1016/j.jvcir.2018.09.019
DO - 10.1016/j.jvcir.2018.09.019
M3 - Article
AN - SCOPUS:85054274833
VL - 56
SP - 256
EP - 264
JO - Journal of Visual Communication and Image Representation
JF - Journal of Visual Communication and Image Representation
SN - 1047-3203
ER -