Scipediacontent: Scipediacontent moved page Draft Content 988305824 to Rukhovich et al 2019a

2021-02-01T18:34:37Z

Scipediacontent moved page Draft Content 988305824 to Rukhovich et al 2019a

Scipediacontent: Created page with " == Abstract == This paper addresses the problem of scale estimation in monocular SLAM by estimating absolute distances between camera centers of consecutive image frames. Th..."

2021-02-01T18:34:33Z

Created page with " == Abstract == This paper addresses the problem of scale estimation in monocular SLAM by estimating absolute distances between camera centers of consecutive image frames. Th..."

New page

== Abstract ==

This paper addresses the problem of scale estimation in monocular SLAM by estimating absolute distances between camera centers of consecutive image frames. These estimates would improve the overall performance of classical (not deep) SLAM systems and allow metric feature locations to be recovered from a single monocular camera. We propose several network architectures that lead to an improvement of scale estimation accuracy over the state of the art. In addition, we exploit a possibility to train the neural network only with synthetic data derived from a computer graphics simulator. Our key insight is that, using only synthetic training inputs, we can achieve similar scale estimation accuracy as that obtained from real data. This fact indicates that fully annotated simulated data is a viable alternative to existing deep-learning-based SLAM systems trained on real (unlabeled) data. Our experiments with unsupervised domain adaptation also show that the difference in visual appearance between simulated and real data does not affect scale estimation results. Our method operates with low-resolution images (0.03 MP), which makes it practical for real-time SLAM applications with a monocular camera.

== Original document ==

The different versions of the original document can be found in:

* [http://arxiv.org/abs/1909.00713 http://arxiv.org/abs/1909.00713]

* [http://arxiv.org/pdf/1909.00713 http://arxiv.org/pdf/1909.00713]

* [http://xplorestaging.ieee.org/ielx7/8982559/9021948/09022554.pdf?arnumber=9022554 http://xplorestaging.ieee.org/ielx7/8982559/9021948/09022554.pdf?arnumber=9022554],
: [http://dx.doi.org/10.1109/iccvw.2019.00108 http://dx.doi.org/10.1109/iccvw.2019.00108]

* [https://dblp.uni-trier.de/db/journals/corr/corr1909.html#abs-1909-00713 https://dblp.uni-trier.de/db/journals/corr/corr1909.html#abs-1909-00713],
: [http://openaccess.thecvf.com/content_ICCVW_2019/papers/CVRSUAD/Rukhovich_Estimation_of_Absolute_Scale_in_Monocular_SLAM_Using_Synthetic_Data_ICCVW_2019_paper.pdf http://openaccess.thecvf.com/content_ICCVW_2019/papers/CVRSUAD/Rukhovich_Estimation_of_Absolute_Scale_in_Monocular_SLAM_Using_Synthetic_Data_ICCVW_2019_paper.pdf],
: [http://ui.adsabs.harvard.edu/abs/2019arXiv190900713R/abstract http://ui.adsabs.harvard.edu/abs/2019arXiv190900713R/abstract],
: [http://openaccess.thecvf.com/content_ICCVW_2019/html/CVRSUAD/Rukhovich_Estimation_of_Absolute_Scale_in_Monocular_SLAM_Using_Synthetic_Data_ICCVW_2019_paper.html http://openaccess.thecvf.com/content_ICCVW_2019/html/CVRSUAD/Rukhovich_Estimation_of_Absolute_Scale_in_Monocular_SLAM_Using_Synthetic_Data_ICCVW_2019_paper.html],
: [https://academic.microsoft.com/#/detail/2997718684 https://academic.microsoft.com/#/detail/2997718684]

← Older revision	Revision as of 18:34, 1 February 2021
(No difference)

Rukhovich et al 2019a - Revision history

Scipediacontent: Scipediacontent moved page Draft Content 988305824 to Rukhovich et al 2019a

Scipediacontent: Created page with " == Abstract == This paper addresses the problem of scale estimation in monocular SLAM by estimating absolute distances between camera centers of consecutive image frames. Th..."