A Holistic Visual Place Recognition Approach Using Lightweight CNNs for Significant ViewPoint and Appearance Changes
- Submitting institution
-
The University of Essex
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 1300
- Type
- D - Journal article
- DOI
-
10.1109/tro.2019.2956352
- Title of journal
- IEEE Transactions on Robotics
- Article number
- -
- First page
- 561
- Volume
- 36
- Issue
- 2
- ISSN
- 1552-3098
- Open access status
- Compliant
- Month of publication
- December
- 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
-
4
- Research group(s)
-
D - Robotics and Embedded Systems (RES)
- Citation count
- 5
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This highly-cited paper presents a lightweight visual place recognition approach, capable of achieving high performance with low computational cost, and feasible for use in mobile robotics under significant viewpoint and appearance changes, the 2 principal challenges within the research topic. Significantly, the results on several benchmark datasets confirm a mean enhancement of 13% in accuracy, and 12x average speedup relative to state-of-the-art methods. This is collaborative work with the Australian Centre for Robotic Vision and ETH Zurich. This paper is a result from our EPSRC National Centre for Nuclear Robotics project (EP/R02572X/1-GBP 11.4M) and underpins ongoing development of the ideas.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -