publication . Preprint . 2017

Towards Stable Adversarial Feature Learning for LiDAR based Loop Closure Detection

Xu, Lingyun; Yin, Peng; Luo, Haibo; Liu, Yunhui; Han, Jianda;
Open Access English
  • Published: 21 Nov 2017
Abstract
Stable feature extraction is the key for the Loop closure detection (LCD) task in the simultaneously localization and mapping (SLAM) framework. In our paper, the feature extraction is operated by using a generative adversarial networks (GANs) based unsupervised learning. GANs are powerful generative models, however, GANs based adversarial learning suffers from training instability. We find that the data-code joint distribution in the adversarial learning is a more complex manifold than in the original GANs. And the loss function that drive the attractive force between synthesis and target distributions is unable for efficient latent code learning for LCD task. T...
Read more
Subjects
free text keywords: Computer Science - Robotics
Related Organizations
View more
Download from
arXiv.org e-Print Archive
Preprint . 2017
Provider: arXiv.org e-Print Archive
25 references, page 1 of 2

[1] R. Mur-Artal, J.MM Montiel, and J. D. Tardos, "Orb-slam: a versatile and accurate monocular slam system," IEEE Transactions on Robotics, Vol.31, pp.1147-1163, 2015. [OpenAIRE]

[2] J. Engel, T. Schops, and D. Cremers, "LSD-SLAM: Large-Scale Direct monocular SLAM," presented at European Conference on Computer Vision, Zurich, Switzerland, 2014. [OpenAIRE]

[3] Choset, Howie, and Keiji Nagatani. "Topological simultaneous localization and mapping (SLAM): toward exact localization without explicit localization." IEEE Transactions on robotics and automation, Vol.17, pp.125-137, 2001. [OpenAIRE]

[4] N. Sünderhauf, P. Neubert, and P. Protzel, "Are we there yet? Challenging SeqSLAM on a 3000 km journey across all four seasons," presented at IEEE International Conference on Robotics and Automation, Karlsruhe, Germany, 2013.

[5] G. Wyeth and M. Milford, "Towards lifelong navigation and mapping in an office environment," Robotics Research, Vol.70, pp.589-603, 2011. [OpenAIRE]

M. Cummins and P. Newman, "Appearance-only SLAM at large scale with FAB-MAP 2.0," International Journal of Robotics Research, Vol.30, pp.1100-1123, 2011.

M. Milford, G. Wyeth, and D. Prasser, "RatSLAM: a hippocampal model for simultaneous localization and mapping," presented at IEEE International Conference on Robotics and Automation, New Orleans, United States, 2004. [OpenAIRE]

M. J. Milford and G. F. Wyeth, "SeqSLAM: Visual route-based navigation for sunny summer days and stormy winter nights," presented at IEEE International Conferece on Robotics and Automation, Minnesota, United States, 2012. [OpenAIRE]

P. C. Ng and S. Henikoff, "SIFT: Predicting amino acid changes that affect protein function," Nucleic Acids Research. Vol.31, pp.3812-3814, 2003. [OpenAIRE]

[10] H. Bay, T. Tuytelaars, and L. Van Gool, "SURF: Speeded up robust features," presented at European Conference on Computer Vision, Graz, Austria, 2006. [OpenAIRE]

[11] E. Rublee, V. Rabaud, K. Konolige, and G. Bradski, "ORB: An efficient alternative to SIFT or SURF," presented at IEEE International Conference on Computer Vision, Barcelona, Spain, 2011. [OpenAIRE]

[12] A. Oliva, "Gist of the scene," Neurobiology of Attention, pp. 251-256, 2005.

[13] N. Sunderhauf, S. Shirazi, F. Dayoub, B. Upcroft, and M. Milford, "On the performance of ConvNet features for place recognition," presented at IEEE Conference on Intelligent Robot and Systems, Germany, 2015. [OpenAIRE]

[14] S. Lowry and M. J. Milford, "Change Removal : Robust Online Learning for Changing Appearance and Changing Viewpoint ," presented at IEEE International Conference on Robotics and Automation, Seattle, United States, 2015.

[15] Z. Chen, A. Jacobson, N. Sünderhauf, B. Upcroft, L. Liu, C. Shen, and I. Reid, "Deep Learning Features at Scale for Visual Place Recognition," presented at IEEE International Automation, Singapore, 2017. [OpenAIRE]

25 references, page 1 of 2
Powered by OpenAIRE Research Graph
Any information missing or wrong?Report an Issue