Deep Manifold Embedding for 3D Object Pose Estimation

Hiroshi Ninomiya, Yasutomo Kawanishi, Daisuke Deguchi, Ichiro Ide, Hiroshi Murase, Norimasa Kobori, Yusuke Nakano

2017

Abstract

Recently, 3D object pose estimation is being focused. The Parametric Eigenspace method is known as one of the fundamental methods for this. It represents the appearance change of an object caused by pose change with a manifold embedded in a low-dimensional subspace. It obtains features by Principal Component Analysis (PCA), which maximizes the appearance variation. However, there is a problem that it cannot handle a pose change with slight appearance change since there is not always a correlation between pose change and appearance change. In this paper, we propose a method that introduces “Deep Manifold Embedding” which maximizes the pose variation directly. We construct a manifold from features extracted from Deep Convolutional Neural Networks (DCNNs) trained with pose information. Pose estimation with the proposed method achieved the best accuracy in experiments using a public dataset.

Paper Citation

in Harvard Style

Ninomiya H., Kawanishi Y., Deguchi D., Ide I., Murase H., Kobori N. and Nakano Y. (2017). Deep Manifold Embedding for 3D Object Pose Estimation . In Proceedings of the 12th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 5: VISAPP, (VISIGRAPP 2017) ISBN 978-989-758-226-4, pages 173-178. DOI: 10.5220/0006101201730178

in Bibtex Style

@conference{visapp17,
author={Hiroshi Ninomiya and Yasutomo Kawanishi and Daisuke Deguchi and Ichiro Ide and Hiroshi Murase and Norimasa Kobori and Yusuke Nakano},
title={Deep Manifold Embedding for 3D Object Pose Estimation},
booktitle={Proceedings of the 12th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 5: VISAPP, (VISIGRAPP 2017)},
year={2017},
pages={173-178},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0006101201730178},
isbn={978-989-758-226-4},
}

in EndNote Style

TY - CONF
JO - Proceedings of the 12th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 5: VISAPP, (VISIGRAPP 2017)
TI - Deep Manifold Embedding for 3D Object Pose Estimation
SN - 978-989-758-226-4
AU - Ninomiya H.
AU - Kawanishi Y.
AU - Deguchi D.
AU - Ide I.
AU - Murase H.
AU - Kobori N.
AU - Nakano Y.
PY - 2017
SP - 173
EP - 178
DO - 10.5220/0006101201730178