An evaluation of deep learning in loop closure detection for visual SLAM

Yifan Xia; Jie Li; Lin Qi; Hui Yu; Junyu Dong

doi:10.1109/iThings-GreenCom-CPSCom-SmartData.2017.18

An evaluation of deep learning in loop closure detection for visual SLAM

Yifan Xia, Jie Li, Lin Qi, Hui Yu, Junyu Dong

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Loop closure detection is a crucial module in simultaneous localization and mapping (SLAM), which reduces the accumulative error in building the environment map. Traditional appearance-based methods mostly utilize hand-crafted features, which are designed based on human expertise. Recent advances in deep learning inspire us to investigate its application in loop closure detection. Different from traditional approaches, deep learning methods automatically learn features from raw data and has better adaptability to complex environment changes. In this paper, we perform a comparison and analysis of several popular deep neural networks and traditional methods for loop closure detection. We evaluate their performance on two open datasets in terms of accuracy and processing time. According to the experimental results, we conclude that deep neural network is suitable for loop closure detection.

Original language	English
Title of host publication	2017 IEEE International Conference on Internet of Things (iThings) and IEEE Green Computing and Communications (GreenCom) and IEEE Cyber, Physical and Social Computing (CPSCom) and IEEE Smart Data (SmartData)
Editors	Geyong Min, Xiaolong Jin, Laurence T. Yang, Yulei Wu, Nektarios Georgalas, Ahmed Al-Dubi
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	85-91
Number of pages	7
ISBN (Electronic)	978-1-5386-3066-2
ISBN (Print)	978-1-5386-3067-9
DOIs	https://doi.org/10.1109/iThings-GreenCom-CPSCom-SmartData.2017.18
Publication status	Published - 1 Feb 2018
Event	Joint 10th IEEE International Conference on Internet of Things, iThings 2017, 13th IEEE International Conference on Green Computing and Communications, GreenCom 2017, 10th IEEE International Conference on Cyber, Physical and Social Computing, CPSCom 2017 and the 3rd IEEE International Conference on Smart Data, Smart Data 2017 - Exeter, United Kingdom Duration: 21 Jun 2017 → 23 Jun 2017

Conference

Conference	Joint 10th IEEE International Conference on Internet of Things, iThings 2017, 13th IEEE International Conference on Green Computing and Communications, GreenCom 2017, 10th IEEE International Conference on Cyber, Physical and Social Computing, CPSCom 2017 and the 3rd IEEE International Conference on Smart Data, Smart Data 2017
Country/Territory	United Kingdom
City	Exeter
Period	21/06/17 → 23/06/17

Keywords

Deep Learning
Loop Closure Detection
Simultaneous Localization and Mapping
noissn

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1109/iThings-GreenCom-CPSCom-SmartData.2017.18

Cite this

Xia, Y., Li, J., Qi, L., Yu, H., & Dong, J. (2018). An evaluation of deep learning in loop closure detection for visual SLAM. In G. Min, X. Jin, L. T. Yang, Y. Wu, N. Georgalas, & A. Al-Dubi (Eds.), 2017 IEEE International Conference on Internet of Things (iThings) and IEEE Green Computing and Communications (GreenCom) and IEEE Cyber, Physical and Social Computing (CPSCom) and IEEE Smart Data (SmartData) (pp. 85-91). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/iThings-GreenCom-CPSCom-SmartData.2017.18

Xia, Yifan ; Li, Jie ; Qi, Lin et al. / An evaluation of deep learning in loop closure detection for visual SLAM. 2017 IEEE International Conference on Internet of Things (iThings) and IEEE Green Computing and Communications (GreenCom) and IEEE Cyber, Physical and Social Computing (CPSCom) and IEEE Smart Data (SmartData). editor / Geyong Min ; Xiaolong Jin ; Laurence T. Yang ; Yulei Wu ; Nektarios Georgalas ; Ahmed Al-Dubi. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 85-91

@inproceedings{d0e6edb7bc374a3d999da2dd0c3aed49,

title = "An evaluation of deep learning in loop closure detection for visual SLAM",

abstract = "Loop closure detection is a crucial module in simultaneous localization and mapping (SLAM), which reduces the accumulative error in building the environment map. Traditional appearance-based methods mostly utilize hand-crafted features, which are designed based on human expertise. Recent advances in deep learning inspire us to investigate its application in loop closure detection. Different from traditional approaches, deep learning methods automatically learn features from raw data and has better adaptability to complex environment changes. In this paper, we perform a comparison and analysis of several popular deep neural networks and traditional methods for loop closure detection. We evaluate their performance on two open datasets in terms of accuracy and processing time. According to the experimental results, we conclude that deep neural network is suitable for loop closure detection.",

keywords = "Deep Learning, Loop Closure Detection, Simultaneous Localization and Mapping, noissn",

author = "Yifan Xia and Jie Li and Lin Qi and Hui Yu and Junyu Dong",

note = "No ISSN found.; Joint 10th IEEE International Conference on Internet of Things, iThings 2017, 13th IEEE International Conference on Green Computing and Communications, GreenCom 2017, 10th IEEE International Conference on Cyber, Physical and Social Computing, CPSCom 2017 and the 3rd IEEE International Conference on Smart Data, Smart Data 2017 ; Conference date: 21-06-2017 Through 23-06-2017",

year = "2018",

month = feb,

day = "1",

doi = "10.1109/iThings-GreenCom-CPSCom-SmartData.2017.18",

language = "English",

isbn = "978-1-5386-3067-9",

pages = "85--91",

editor = "Geyong Min and Xiaolong Jin and Yang, {Laurence T.} and Yulei Wu and Nektarios Georgalas and Ahmed Al-Dubi",

booktitle = "2017 IEEE International Conference on Internet of Things (iThings) and IEEE Green Computing and Communications (GreenCom) and IEEE Cyber, Physical and Social Computing (CPSCom) and IEEE Smart Data (SmartData)",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

address = "United States",

}

Xia, Y, Li, J, Qi, L, Yu, H & Dong, J 2018, An evaluation of deep learning in loop closure detection for visual SLAM. in G Min, X Jin, LT Yang, Y Wu, N Georgalas & A Al-Dubi (eds), 2017 IEEE International Conference on Internet of Things (iThings) and IEEE Green Computing and Communications (GreenCom) and IEEE Cyber, Physical and Social Computing (CPSCom) and IEEE Smart Data (SmartData). Institute of Electrical and Electronics Engineers Inc., pp. 85-91, Joint 10th IEEE International Conference on Internet of Things, iThings 2017, 13th IEEE International Conference on Green Computing and Communications, GreenCom 2017, 10th IEEE International Conference on Cyber, Physical and Social Computing, CPSCom 2017 and the 3rd IEEE International Conference on Smart Data, Smart Data 2017, Exeter, United Kingdom, 21/06/17. https://doi.org/10.1109/iThings-GreenCom-CPSCom-SmartData.2017.18

An evaluation of deep learning in loop closure detection for visual SLAM. / Xia, Yifan; Li, Jie; Qi, Lin et al.
2017 IEEE International Conference on Internet of Things (iThings) and IEEE Green Computing and Communications (GreenCom) and IEEE Cyber, Physical and Social Computing (CPSCom) and IEEE Smart Data (SmartData). ed. / Geyong Min; Xiaolong Jin; Laurence T. Yang; Yulei Wu; Nektarios Georgalas; Ahmed Al-Dubi. Institute of Electrical and Electronics Engineers Inc., 2018. p. 85-91.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - An evaluation of deep learning in loop closure detection for visual SLAM

AU - Xia, Yifan

AU - Li, Jie

AU - Qi, Lin

AU - Yu, Hui

AU - Dong, Junyu

N1 - No ISSN found.

PY - 2018/2/1

Y1 - 2018/2/1

N2 - Loop closure detection is a crucial module in simultaneous localization and mapping (SLAM), which reduces the accumulative error in building the environment map. Traditional appearance-based methods mostly utilize hand-crafted features, which are designed based on human expertise. Recent advances in deep learning inspire us to investigate its application in loop closure detection. Different from traditional approaches, deep learning methods automatically learn features from raw data and has better adaptability to complex environment changes. In this paper, we perform a comparison and analysis of several popular deep neural networks and traditional methods for loop closure detection. We evaluate their performance on two open datasets in terms of accuracy and processing time. According to the experimental results, we conclude that deep neural network is suitable for loop closure detection.

AB - Loop closure detection is a crucial module in simultaneous localization and mapping (SLAM), which reduces the accumulative error in building the environment map. Traditional appearance-based methods mostly utilize hand-crafted features, which are designed based on human expertise. Recent advances in deep learning inspire us to investigate its application in loop closure detection. Different from traditional approaches, deep learning methods automatically learn features from raw data and has better adaptability to complex environment changes. In this paper, we perform a comparison and analysis of several popular deep neural networks and traditional methods for loop closure detection. We evaluate their performance on two open datasets in terms of accuracy and processing time. According to the experimental results, we conclude that deep neural network is suitable for loop closure detection.

KW - Deep Learning

KW - Loop Closure Detection

KW - Simultaneous Localization and Mapping

KW - noissn

UR - http://www.scopus.com/inward/record.url?scp=85047323965&partnerID=8YFLogxK

U2 - 10.1109/iThings-GreenCom-CPSCom-SmartData.2017.18

DO - 10.1109/iThings-GreenCom-CPSCom-SmartData.2017.18

M3 - Conference contribution

AN - SCOPUS:85047323965

SN - 978-1-5386-3067-9

SP - 85

EP - 91

BT - 2017 IEEE International Conference on Internet of Things (iThings) and IEEE Green Computing and Communications (GreenCom) and IEEE Cyber, Physical and Social Computing (CPSCom) and IEEE Smart Data (SmartData)

A2 - Min, Geyong

A2 - Jin, Xiaolong

A2 - Yang, Laurence T.

A2 - Wu, Yulei

A2 - Georgalas, Nektarios

A2 - Al-Dubi, Ahmed

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - Joint 10th IEEE International Conference on Internet of Things, iThings 2017, 13th IEEE International Conference on Green Computing and Communications, GreenCom 2017, 10th IEEE International Conference on Cyber, Physical and Social Computing, CPSCom 2017 and the 3rd IEEE International Conference on Smart Data, Smart Data 2017

Y2 - 21 June 2017 through 23 June 2017

ER -

Xia Y, Li J, Qi L, Yu H, Dong J. An evaluation of deep learning in loop closure detection for visual SLAM. In Min G, Jin X, Yang LT, Wu Y, Georgalas N, Al-Dubi A, editors, 2017 IEEE International Conference on Internet of Things (iThings) and IEEE Green Computing and Communications (GreenCom) and IEEE Cyber, Physical and Social Computing (CPSCom) and IEEE Smart Data (SmartData). Institute of Electrical and Electronics Engineers Inc. 2018. p. 85-91 doi: 10.1109/iThings-GreenCom-CPSCom-SmartData.2017.18

An evaluation of deep learning in loop closure detection for visual SLAM

Abstract

Conference

Keywords

UN SDGs

Access to Document

Fingerprint

Cite this