NICE-SLAM: Neural Implicit Scalable Encoding for SLAM

Zihan Zhu; Songyou Peng; Viktor Larsson; Weiwei Xu; Hujun Bao; Zhaopeng Cui; Martin R Oswald; Marc Pollefeys

doi:10.1109/CVPR52688.2022.01245

NICE-SLAM: Neural Implicit Scalable Encoding for SLAM

Zihan Zhu, Songyou Peng, Viktor Larsson, Weiwei Xu, Hujun Bao, Zhaopeng Cui, Martin R Oswald, Marc Pollefeys

Research output: Chapter in Book/Report/Conference proceeding › Paper in conference proceeding › peer-review

Abstract

Neural implicit representations have recently shown encouraging results in various domains, including promising progress in simultaneous localization and mapping (SLAM). Nevertheless, existing methods produce over- smoothed scene reconstructions and have difficulty scaling up to large scenes. These limitations are mainly due to their simple fully-connected network architecture that does not incorporate local information in the observations. In this paper, we present NICE-SLAM, a dense SLAM system that incorporates multi-level local information by introducing a hierarchical scene representation. Optimizing this representation with pre-trained geometric priors enables detailed reconstruction on large indoor scenes. Compared to recent neural implicit SLAM systems, our approach is more scalable, efficient, and robust. Experiments on five challenging datasets demonstrate competitive results of NICE-SLAM in both mapping and tracking quality.

Original language	English
Title of host publication	Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition
Publisher	IEEE - Institute of Electrical and Electronics Engineers Inc.
Pages	12786-12796
Number of pages	11
ISBN (Electronic)	978-1-6654-6946-3
ISBN (Print)	978-1-6654-6947-0
DOIs	https://doi.org/10.1109/CVPR52688.2022.01245
Publication status	Published - 2022
Event	2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022 - New Orleans, United States Duration: 2022 Jun 19 → 2022 Jun 24

Conference

Conference	2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022
Country/Territory	United States
City	New Orleans
Period	2022/06/19 → 2022/06/24

Subject classification (UKÄ)

Computer graphics and computer vision

Access to Document

10.1109/CVPR52688.2022.01245

Cite this

@inproceedings{1a9abc5b532840e19b084d990f417366,

title = "NICE-SLAM: Neural Implicit Scalable Encoding for SLAM",

abstract = "Neural implicit representations have recently shown encouraging results in various domains, including promising progress in simultaneous localization and mapping (SLAM). Nevertheless, existing methods produce over- smoothed scene reconstructions and have difficulty scaling up to large scenes. These limitations are mainly due to their simple fully-connected network architecture that does not incorporate local information in the observations. In this paper, we present NICE-SLAM, a dense SLAM system that incorporates multi-level local information by introducing a hierarchical scene representation. Optimizing this representation with pre-trained geometric priors enables detailed reconstruction on large indoor scenes. Compared to recent neural implicit SLAM systems, our approach is more scalable, efficient, and robust. Experiments on five challenging datasets demonstrate competitive results of NICE-SLAM in both mapping and tracking quality. ",

author = "Zihan Zhu and Songyou Peng and Viktor Larsson and Weiwei Xu and Hujun Bao and Zhaopeng Cui and Oswald, {Martin R} and Marc Pollefeys",

year = "2022",

doi = "10.1109/CVPR52688.2022.01245",

language = "English",

isbn = "978-1-6654-6947-0",

pages = "12786--12796",

booktitle = "Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition",

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

address = "United States",

note = "2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022 ; Conference date: 19-06-2022 Through 24-06-2022",

}

Zhu, Z, Peng, S, Larsson, V, Xu, W, Bao, H, Cui, Z, Oswald, MR & Pollefeys, M 2022, NICE-SLAM: Neural Implicit Scalable Encoding for SLAM. in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE - Institute of Electrical and Electronics Engineers Inc., pp. 12786-12796, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022, New Orleans, Louisiana, United States, 2022/06/19. https://doi.org/10.1109/CVPR52688.2022.01245

NICE-SLAM: Neural Implicit Scalable Encoding for SLAM. / Zhu, Zihan; Peng, Songyou; Larsson, Viktor et al.
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE - Institute of Electrical and Electronics Engineers Inc., 2022. p. 12786-12796.

Research output: Chapter in Book/Report/Conference proceeding › Paper in conference proceeding › peer-review

TY - GEN

T1 - NICE-SLAM: Neural Implicit Scalable Encoding for SLAM

AU - Zhu, Zihan

AU - Peng, Songyou

AU - Larsson, Viktor

AU - Xu, Weiwei

AU - Bao, Hujun

AU - Cui, Zhaopeng

AU - Oswald, Martin R

AU - Pollefeys, Marc

PY - 2022

Y1 - 2022

N2 - Neural implicit representations have recently shown encouraging results in various domains, including promising progress in simultaneous localization and mapping (SLAM). Nevertheless, existing methods produce over- smoothed scene reconstructions and have difficulty scaling up to large scenes. These limitations are mainly due to their simple fully-connected network architecture that does not incorporate local information in the observations. In this paper, we present NICE-SLAM, a dense SLAM system that incorporates multi-level local information by introducing a hierarchical scene representation. Optimizing this representation with pre-trained geometric priors enables detailed reconstruction on large indoor scenes. Compared to recent neural implicit SLAM systems, our approach is more scalable, efficient, and robust. Experiments on five challenging datasets demonstrate competitive results of NICE-SLAM in both mapping and tracking quality.

AB - Neural implicit representations have recently shown encouraging results in various domains, including promising progress in simultaneous localization and mapping (SLAM). Nevertheless, existing methods produce over- smoothed scene reconstructions and have difficulty scaling up to large scenes. These limitations are mainly due to their simple fully-connected network architecture that does not incorporate local information in the observations. In this paper, we present NICE-SLAM, a dense SLAM system that incorporates multi-level local information by introducing a hierarchical scene representation. Optimizing this representation with pre-trained geometric priors enables detailed reconstruction on large indoor scenes. Compared to recent neural implicit SLAM systems, our approach is more scalable, efficient, and robust. Experiments on five challenging datasets demonstrate competitive results of NICE-SLAM in both mapping and tracking quality.

UR - https://pengsongyou.github.io/nice-slam

U2 - 10.1109/CVPR52688.2022.01245

DO - 10.1109/CVPR52688.2022.01245

M3 - Paper in conference proceeding

SN - 978-1-6654-6947-0

SP - 12786

EP - 12796

BT - Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

PB - IEEE - Institute of Electrical and Electronics Engineers Inc.

T2 - 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022

Y2 - 19 June 2022 through 24 June 2022

ER -

NICE-SLAM: Neural Implicit Scalable Encoding for SLAM

Abstract

Conference

Subject classification (UKÄ)

Access to Document

Other files and links

Fingerprint

Cite this