When Regression Meets Manifold Learning for Object Recognition and Pose Estimation

Bui M, Zakharov S, Albarqouni S, Ilic S, Navab N (2018)

Publication Type: Conference contribution

Publication year: 2018

Journal

Proceedings - IEEE International Conference on Robotics and Automation Institute of Electrical and Electronics Engineers

Publisher: Institute of Electrical and Electronics Engineers Inc.

Pages Range: 6140-6146

Conference Proceedings Title: Proceedings - IEEE International Conference on Robotics and Automation

Event location: Brisbane, QLD, AUS

ISBN: 9781538630815

DOI: 10.1109/ICRA.2018.8460654

Abstract

In this work, we propose a method for object recognition and pose estimation from depth images using convolutional neural networks. Previous methods addressing this problem rely on manifold learning to learn low dimensional viewpoint descriptors and employ them in a nearest neighbor search on an estimated descriptor space. In comparison we create an efficient multi-task learning framework combining manifold descriptor learning and pose regression. By combining the strengths of manifold learning using triplet loss and pose regression, we could either estimate the pose directly reducing the complexity compared to NN search, or use the learned descriptor for the NN descriptor matching. By in depth experimental evaluation of the novel loss function we observed that the view descriptors learned by the network are much more discriminative resulting in almost 30% increase regarding relative pose accuracy compared to related works. On the other hand, regarding directly regressed poses we obtained important improvement compared to simple pose regression. By leveraging the advantages of both manifold learning and regression tasks, we are able to improve the current state-of-the-art for object recognition and pose retrieval.

Involved external institutions

Technische Universität München (TUM) DE Germany (DE)

How to cite

APA:

Bui, M., Zakharov, S., Albarqouni, S., Ilic, S., & Navab, N. (2018). When Regression Meets Manifold Learning for Object Recognition and Pose Estimation. In Proceedings - IEEE International Conference on Robotics and Automation (pp. 6140-6146). Brisbane, QLD, AUS: Institute of Electrical and Electronics Engineers Inc..

MLA:

Bui, Mai, et al. "When Regression Meets Manifold Learning for Object Recognition and Pose Estimation." Proceedings of the 2018 IEEE International Conference on Robotics and Automation, ICRA 2018, Brisbane, QLD, AUS Institute of Electrical and Electronics Engineers Inc., 2018. 6140-6146.

BibTeX: Download