Slice Transformer and Self-supervised Learning for 6DoF Localization in 3D Point Cloud Maps

Muhammad Ibrahim; Naveed Akhtar; Saeed Anwar; Michael Wise; Ajmal Mian

doi:10.1109/ICRA48891.2023.10161128

Slice Transformer and Self-supervised Learning for 6DoF Localization in 3D Point Cloud Maps

Muhammad Ibrahim, Naveed Akhtar, Saeed Anwar, Michael Wise, Ajmal Mian

Research output: Chapter in Book/Conference paper › Conference paper › peer-review

Abstract

Precise localization is critical for autonomous vehicles. We present a self-supervised learning method that employs transformers for the first time for the task of outdoor localization using LiDAR data. We propose a pre-text task that reorganizes the slices of a 360° LiDAR scan to leverage its axial properties. Our model, called Slice Transformer, employs multi-head attention while systematically processing the slices. To the best of our knowledge, this is the first instance of leveraging multi-head attention for outdoor point clouds. We additionally introduce the Perth-Wadataset, which provides a large-scale LiDAR map of Perth city in Western Australia, covering 4km2area. Localization annotations are provided for Perth - Wa.The proposed localization method is thoroughly evaluated on Perth-WA and Appollo-SouthBay datasets. We also establish the efficacy of our self-supervised learning approach for the common downstream task of object classification using ModelNet40 and ScanNN datasets. The code and Perth-WA data will be publicly released.

Original language	English
Title of host publication	Proceedings - ICRA 2023
Subtitle of host publication	IEEE International Conference on Robotics and Automation
Publisher	IEEE, Institute of Electrical and Electronics Engineers
Pages	11763-11770
Number of pages	8
ISBN (Electronic)	9798350323658
DOIs	https://doi.org/10.1109/ICRA48891.2023.10161128
Publication status	Published - 2023
Event	2023 IEEE International Conference on Robotics and Automation - London, United Kingdom Duration: 29 May 2023 → 2 Jun 2023

Publication series

Name	Proceedings - IEEE International Conference on Robotics and Automation
Volume	2023-May
ISSN (Print)	1050-4729

Conference

Conference	2023 IEEE International Conference on Robotics and Automation
Abbreviated title	ICRA 2023
Country/Territory	United Kingdom
City	London
Period	29/05/23 → 2/06/23

Access to Document

10.1109/ICRA48891.2023.10161128

https://arxiv.org/abs/2301.08957v2Licence: CC BY

Cite this

Ibrahim, M., Akhtar, N., Anwar, S., Wise, M., & Mian, A. (2023). Slice Transformer and Self-supervised Learning for 6DoF Localization in 3D Point Cloud Maps. In Proceedings - ICRA 2023: IEEE International Conference on Robotics and Automation (pp. 11763-11770). (Proceedings - IEEE International Conference on Robotics and Automation; Vol. 2023-May). IEEE, Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/ICRA48891.2023.10161128

Ibrahim, Muhammad ; Akhtar, Naveed ; Anwar, Saeed et al. / Slice Transformer and Self-supervised Learning for 6DoF Localization in 3D Point Cloud Maps. Proceedings - ICRA 2023: IEEE International Conference on Robotics and Automation. IEEE, Institute of Electrical and Electronics Engineers, 2023. pp. 11763-11770 (Proceedings - IEEE International Conference on Robotics and Automation).

@inproceedings{ca6a86527ae34d6fb8776afcb95cb058,

title = "Slice Transformer and Self-supervised Learning for 6DoF Localization in 3D Point Cloud Maps",

abstract = "Precise localization is critical for autonomous vehicles. We present a self-supervised learning method that employs transformers for the first time for the task of outdoor localization using LiDAR data. We propose a pre-text task that reorganizes the slices of a 360° LiDAR scan to leverage its axial properties. Our model, called Slice Transformer, employs multi-head attention while systematically processing the slices. To the best of our knowledge, this is the first instance of leveraging multi-head attention for outdoor point clouds. We additionally introduce the Perth-Wadataset, which provides a large-scale LiDAR map of Perth city in Western Australia, covering 4km2area. Localization annotations are provided for Perth - Wa.The proposed localization method is thoroughly evaluated on Perth-WA and Appollo-SouthBay datasets. We also establish the efficacy of our self-supervised learning approach for the common downstream task of object classification using ModelNet40 and ScanNN datasets. The code and Perth-WA data will be publicly released.",

author = "Muhammad Ibrahim and Naveed Akhtar and Saeed Anwar and Michael Wise and Ajmal Mian",

note = "Funding Information: Professor Ajmal Mian is the recipient of an Australian Research Council Future Fellowship Award (project number FT210100268) funded by the Australian Government. Dr. Naveed Akhtar is recipient of an Office of National Intelligence National Intelligence Postdoctoral Grant (project number NIPG-2021-001) funded by the Australian Government. Publisher Copyright: {\textcopyright} 2023 IEEE.; 2023 IEEE International Conference on Robotics and Automation, ICRA 2023 ; Conference date: 29-05-2023 Through 02-06-2023",

year = "2023",

doi = "10.1109/ICRA48891.2023.10161128",

language = "English",

series = "Proceedings - IEEE International Conference on Robotics and Automation",

publisher = "IEEE, Institute of Electrical and Electronics Engineers",

pages = "11763--11770",

booktitle = "Proceedings - ICRA 2023",

address = "United States",

}

Ibrahim, M, Akhtar, N, Anwar, S, Wise, M & Mian, A 2023, Slice Transformer and Self-supervised Learning for 6DoF Localization in 3D Point Cloud Maps. in Proceedings - ICRA 2023: IEEE International Conference on Robotics and Automation. Proceedings - IEEE International Conference on Robotics and Automation, vol. 2023-May, IEEE, Institute of Electrical and Electronics Engineers, pp. 11763-11770, 2023 IEEE International Conference on Robotics and Automation, London, United Kingdom, 29/05/23. https://doi.org/10.1109/ICRA48891.2023.10161128

Slice Transformer and Self-supervised Learning for 6DoF Localization in 3D Point Cloud Maps. / Ibrahim, Muhammad; Akhtar, Naveed; Anwar, Saeed et al.
Proceedings - ICRA 2023: IEEE International Conference on Robotics and Automation. IEEE, Institute of Electrical and Electronics Engineers, 2023. p. 11763-11770 (Proceedings - IEEE International Conference on Robotics and Automation; Vol. 2023-May).

Research output: Chapter in Book/Conference paper › Conference paper › peer-review

TY - GEN

T1 - Slice Transformer and Self-supervised Learning for 6DoF Localization in 3D Point Cloud Maps

AU - Ibrahim, Muhammad

AU - Akhtar, Naveed

AU - Anwar, Saeed

AU - Wise, Michael

AU - Mian, Ajmal

N1 - Funding Information: Professor Ajmal Mian is the recipient of an Australian Research Council Future Fellowship Award (project number FT210100268) funded by the Australian Government. Dr. Naveed Akhtar is recipient of an Office of National Intelligence National Intelligence Postdoctoral Grant (project number NIPG-2021-001) funded by the Australian Government. Publisher Copyright: © 2023 IEEE.

PY - 2023

Y1 - 2023

N2 - Precise localization is critical for autonomous vehicles. We present a self-supervised learning method that employs transformers for the first time for the task of outdoor localization using LiDAR data. We propose a pre-text task that reorganizes the slices of a 360° LiDAR scan to leverage its axial properties. Our model, called Slice Transformer, employs multi-head attention while systematically processing the slices. To the best of our knowledge, this is the first instance of leveraging multi-head attention for outdoor point clouds. We additionally introduce the Perth-Wadataset, which provides a large-scale LiDAR map of Perth city in Western Australia, covering 4km2area. Localization annotations are provided for Perth - Wa.The proposed localization method is thoroughly evaluated on Perth-WA and Appollo-SouthBay datasets. We also establish the efficacy of our self-supervised learning approach for the common downstream task of object classification using ModelNet40 and ScanNN datasets. The code and Perth-WA data will be publicly released.

AB - Precise localization is critical for autonomous vehicles. We present a self-supervised learning method that employs transformers for the first time for the task of outdoor localization using LiDAR data. We propose a pre-text task that reorganizes the slices of a 360° LiDAR scan to leverage its axial properties. Our model, called Slice Transformer, employs multi-head attention while systematically processing the slices. To the best of our knowledge, this is the first instance of leveraging multi-head attention for outdoor point clouds. We additionally introduce the Perth-Wadataset, which provides a large-scale LiDAR map of Perth city in Western Australia, covering 4km2area. Localization annotations are provided for Perth - Wa.The proposed localization method is thoroughly evaluated on Perth-WA and Appollo-SouthBay datasets. We also establish the efficacy of our self-supervised learning approach for the common downstream task of object classification using ModelNet40 and ScanNN datasets. The code and Perth-WA data will be publicly released.

UR - http://www.scopus.com/inward/record.url?scp=85168670810&partnerID=8YFLogxK

U2 - 10.1109/ICRA48891.2023.10161128

DO - 10.1109/ICRA48891.2023.10161128

M3 - Conference paper

AN - SCOPUS:85168670810

T3 - Proceedings - IEEE International Conference on Robotics and Automation

SP - 11763

EP - 11770

BT - Proceedings - ICRA 2023

PB - IEEE, Institute of Electrical and Electronics Engineers

T2 - 2023 IEEE International Conference on Robotics and Automation

Y2 - 29 May 2023 through 2 June 2023

ER -

Ibrahim M, Akhtar N, Anwar S, Wise M , Mian A. Slice Transformer and Self-supervised Learning for 6DoF Localization in 3D Point Cloud Maps. In Proceedings - ICRA 2023: IEEE International Conference on Robotics and Automation. IEEE, Institute of Electrical and Electronics Engineers. 2023. p. 11763-11770. (Proceedings - IEEE International Conference on Robotics and Automation). doi: 10.1109/ICRA48891.2023.10161128

Slice Transformer and Self-supervised Learning for 6DoF Localization in 3D Point Cloud Maps

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

3D Scene understanding from LiDAR point clouds

Cite this