Learning Kalman Network: A deep monocular visual odometry for on-road driving
- Submitting institution
-
The University of Sheffield
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 5227
- Type
- D - Journal article
- DOI
-
10.1016/j.robot.2019.07.004
- Title of journal
- Robotics and Autonomous Systems
- Article number
- 103234
- First page
- -
- Volume
- 121
- Issue
- -
- ISSN
- 0921-8890
- Open access status
- Technical exception
- Month of publication
- July
- Year of publication
- 2019
- URL
-
-
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
5
- Research group(s)
-
J - Visual Computing
- Citation count
- 3
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- The paper studies a novel learning-based monocular SLAM that performs simultaneous localisation and mapping using a deep network, applied to mobile robots. The results show that the approach was ranked as the best monocular odometry (in terms of accuracy of localisation) using the most popular KITTI Vision Benchmark Suite. The paper was one of the core achievements of the Horizon 2020 ILIAD project (https://iliad-project.eu/) to deploy forklifts long-term in infrastructure-free warehouses. A video of the work (https://youtu.be/Ccj1O7yndIk) has 10K YouTube views. The preliminary version of this paper was presented at IROS, 2018 (https://arxiv.org/abs/1803.02286).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -