Share Email Print
cover

Proceedings Paper

Towards 3D object recognition with contractive autoencoders
Author(s): Bo Liu; Lingcheng Kong; Jianghai Zhao; Jinghua Wu; Zhiying Tan
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Nowadays, object recognition in 3D scenes has become an emerging challenge with various applications. However, an object can’t be represented well by artificial features derived only from 2D images or depth images separately, and supervised learning method usually requires lots of manually labeled data. To address those limitations, we propose a cross-modality deep learning framework based on contractive autoencoders for 3D scenes object recognition. In particular, we use contractive autoencoding to learn feature representations from 2D and depth images at the same time in an unsupervised way, it is possible to capture their joint information to reinforce detector training. Experiments on 3D image dataset demonstrate the effectiveness of the proposed method for 3D scene object recognition.

Paper Details

Date Published: 29 August 2016
PDF: 7 pages
Proc. SPIE 10033, Eighth International Conference on Digital Image Processing (ICDIP 2016), 100330S (29 August 2016); doi: 10.1117/12.2243988
Show Author Affiliations
Bo Liu, Hefei Institute of Physical Science (China)
Jiangsu Industrial Technology Research Institute (China)
Lingcheng Kong, Hefei Institute of Physical Science (China)
Jiangsu Industrial Technology Research Institute (China)
Jianghai Zhao, Hefei Institute of Physical Science (China)
Jiangsu Industrial Technology Research Institute (China)
Jinghua Wu, Hefei Institute of Physical Science (China)
Jiangsu Industrial Technology Research Institute (China)
Zhiying Tan, Hefei Institute of Physical Science (China)
Jiangsu Industrial Technology Research Institute (China)


Published in SPIE Proceedings Vol. 10033:
Eighth International Conference on Digital Image Processing (ICDIP 2016)
Charles M. Falco; Xudong Jiang, Editor(s)

© SPIE. Terms of Use
Back to Top