Please wait a minute...
FITEE
Front. Inform. Technol. Electron. Eng.  2018, Vol. 19 Issue (11): 1397-1408    DOI:
    
Image-based 3D model retrieval using manifold learning
Pan-pan MU, San-yuan ZHANG , Yin ZHANG , Xiu-zi YE, Xiang PAN
College of Computer Science and Technology, Zhejiang University, Hangzhou 310027, China
College of Mathematics and Information Science, Wenzhou University, Wenzhou 325003, China
College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023, China
Download:   PDF(0KB)
Export: BibTeX | EndNote (RIS)      

Abstract  We propose a new framework for image-based three-dimensional (3D) model retrieval. We first model the query image
as a Euclidean point. Then we model all projected views of a 3D model as a symmetric positive definite (SPD) matrix, which is a
point on a Riemannian manifold. Thus, the image-based 3D model retrieval is reduced to a problem of Euclid-to-Riemann metric
learning. To solve this heterogeneous matching problem, we map the Euclidean space and SPD Riemannian manifold to the same
high-dimensional Hilbert space, thus shrinking the great gap between them. Finally, we design an optimization algorithm to learn
a metric in this Hilbert space using a kernel trick. Any new image descriptors, such as the features from deep learning, can be easily
embedded in our framework. Experimental results show the advantages of our approach over the state-of-the-art  methods for
image-based 3D model retrieval.

Key words： Model retrieval      Euclidean space      Riemannian manifold      Hilbert space      Metric learning     
Received: 01 December 2016      Published: 13 June 2019
Service
E-mail this article
Add to my bookshelf
Add to citation manager
E-mail Alert
RSS
Articles by authors
Pan-pan MU
San-yuan ZHANG
Yin ZHANG
Xiu-zi YE
Xiang PAN
Cite this article:

Pan-pan MU, San-yuan ZHANG , Yin ZHANG , Xiu-zi YE, Xiang PAN. Image-based 3D model retrieval using manifold learning. Front. Inform. Technol. Electron. Eng., 2018, 19(11): 1397-1408.

URL:

http://www.zjujournals.com/xueshu/fitee/     OR     http://www.zjujournals.com/xueshu/fitee/Y2018/V19/I11/1397


Image-based 3D model retrieval using manifold learning

We propose a new framework for image-based three-dimensional (3D) model retrieval. We first model the query image
as a Euclidean point. Then we model all projected views of a 3D model as a symmetric positive definite (SPD) matrix, which is a
point on a Riemannian manifold. Thus, the image-based 3D model retrieval is reduced to a problem of Euclid-to-Riemann metric
learning. To solve this heterogeneous matching problem, we map the Euclidean space and SPD Riemannian manifold to the same
high-dimensional Hilbert space, thus shrinking the great gap between them. Finally, we design an optimization algorithm to learn
a metric in this Hilbert space using a kernel trick. Any new image descriptors, such as the features from deep learning, can be easily
embedded in our framework. Experimental results show the advantages of our approach over the state-of-the-art  methods for
image-based 3D model retrieval.

关键词： Model retrieval,  Euclidean space,  Riemannian manifold,  Hilbert space,  Metric learning 
[1] Hong Shao, Shuang Chen, Jie-yi Zhao, Wen-cheng Cui, Tian-shu Yu. Face recognition based on subset selection via metric learning on manifold[J]. Front. Inform. Technol. Electron. Eng., 2015, 16(12): 1046-1058.