Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

LaMAR: Benchmarking Localization and Mapping for Augmented Reality

Sarlin, Paul-Edouard ; Dusmanu, Mihai ; Schönberger, Johannes L ; Speciale, Pablo ; Gruber, Lukas ; Larsson, Viktor LU ; Miksik, Ondrej and Pollefeys, Marc (2022) 17th European Conference on Computer Vision, ECCV 2022 In Lecture Notes in Computer Science 13667. p.686-704
Abstract
Localization and mapping is the foundational technology for augmented reality (AR) that enables sharing and persistence of digital content in the real world. While significant progress has been made, researchers are still mostly driven by unrealistic benchmarks not representative of real-world AR scenarios. In particular, benchmarks are often based on small-scale datasets with low scene diversity, captured from stationary cameras, and lacking other sensor inputs like inertial, radio, or depth data. Furthermore, ground-truth (GT) accuracy is mostly insufficient to satisfy AR requirements. To close this gap, we introduce a new benchmark with a comprehensive capture and GT pipeline, which allow us to co-register realistic AR trajectories in... (More)
Localization and mapping is the foundational technology for augmented reality (AR) that enables sharing and persistence of digital content in the real world. While significant progress has been made, researchers are still mostly driven by unrealistic benchmarks not representative of real-world AR scenarios. In particular, benchmarks are often based on small-scale datasets with low scene diversity, captured from stationary cameras, and lacking other sensor inputs like inertial, radio, or depth data. Furthermore, ground-truth (GT) accuracy is mostly insufficient to satisfy AR requirements. To close this gap, we introduce a new benchmark with a comprehensive capture and GT pipeline, which allow us to co-register realistic AR trajectories in diverse scenes and from heterogeneous devices at scale. To establish accurate GT, our pipeline robustly aligns the captured trajectories against laser scans in a fully automatic manner. Based on this pipeline, we publish a benchmark dataset of diverse and large-scale scenes recorded with head-mounted and hand-held AR devices. We extend several state-of-the-art methods to take advantage of the AR specific setup and evaluate them on our benchmark. Based on the results, we present novel insights on current research gaps to provide avenues for future work in the community. (Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; ; and
publishing date
type
Chapter in Book/Report/Conference proceeding
publication status
published
subject
host publication
Computer Vision – ECCV 2022 : 7th European Conference, Tel Aviv, Israel, October 23–27, 2022, Proceedings, Part VII - 7th European Conference, Tel Aviv, Israel, October 23–27, 2022, Proceedings, Part VII
series title
Lecture Notes in Computer Science
volume
13667
pages
19 pages
publisher
Springer
conference name
17th European Conference on Computer Vision, ECCV 2022
conference location
Tel Aviv, Israel
conference dates
2022-10-23 - 2022-10-27
external identifiers
  • scopus:85142705973
ISSN
0302-9743
1611-3349
ISBN
978-3-031-20070-0
978-3-031-20071-7
DOI
10.1007/978-3-031-20071-7_40
language
English
LU publication?
no
id
028bc9c7-561f-45ba-9277-5a14932c4d51
date added to LUP
2022-11-25 17:03:10
date last changed
2024-06-25 05:10:53
@inproceedings{028bc9c7-561f-45ba-9277-5a14932c4d51,
  abstract     = {{Localization and mapping is the foundational technology for augmented reality (AR) that enables sharing and persistence of digital content in the real world. While significant progress has been made, researchers are still mostly driven by unrealistic benchmarks not representative of real-world AR scenarios. In particular, benchmarks are often based on small-scale datasets with low scene diversity, captured from stationary cameras, and lacking other sensor inputs like inertial, radio, or depth data. Furthermore, ground-truth (GT) accuracy is mostly insufficient to satisfy AR requirements. To close this gap, we introduce a new benchmark with a comprehensive capture and GT pipeline, which allow us to co-register realistic AR trajectories in diverse scenes and from heterogeneous devices at scale. To establish accurate GT, our pipeline robustly aligns the captured trajectories against laser scans in a fully automatic manner. Based on this pipeline, we publish a benchmark dataset of diverse and large-scale scenes recorded with head-mounted and hand-held AR devices. We extend several state-of-the-art methods to take advantage of the AR specific setup and evaluate them on our benchmark. Based on the results, we present novel insights on current research gaps to provide avenues for future work in the community.}},
  author       = {{Sarlin, Paul-Edouard and Dusmanu, Mihai and Schönberger, Johannes L and Speciale, Pablo and Gruber, Lukas and Larsson, Viktor and Miksik, Ondrej and Pollefeys, Marc}},
  booktitle    = {{Computer Vision – ECCV 2022 : 7th European Conference, Tel Aviv, Israel, October 23–27, 2022, Proceedings, Part VII}},
  isbn         = {{978-3-031-20070-0}},
  issn         = {{0302-9743}},
  language     = {{eng}},
  pages        = {{686--704}},
  publisher    = {{Springer}},
  series       = {{Lecture Notes in Computer Science}},
  title        = {{LaMAR: Benchmarking Localization and Mapping for Augmented Reality}},
  url          = {{http://dx.doi.org/10.1007/978-3-031-20071-7_40}},
  doi          = {{10.1007/978-3-031-20071-7_40}},
  volume       = {{13667}},
  year         = {{2022}},
}