Diffusion Based Robust LiDAR Place Recognition

Benjamin Krummenacher; Jonas Frey; Marco Hutter; Olga Vysotska; Turcan Tuna

REVIEW 4 major objections 4 minor 25 references

Diffusion Based Robust LiDAR Place Recognition

T0 review · 4 major / 4 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A diffusion model trained only on synthetic LiDAR scans can find a robot's global position inside a building.

desk verdict A solid applied robotics paper on diffusion-based LiDAR relocalization against a building mesh; the main weakness is that experiments are scoped to same-mesh, same-time, and Table I has an impossible mean/median pair. read the letter →

arxiv 2504.12412 v1 pith:3VMZXS3K submitted 2025-04-16 cs.RO cs.LG

Benjamin Krummenacher , Jonas Frey , Turcan Tuna , Olga Vysotska , Marco Hutter This is my paper · ORCID

classification cs.ROcs.LG

keywords LiDARplacerecognitiondiffusionmodelgloballocalizationconstructionsitessyntheticpointcloudsperceptualaliasingcloudregistrationmulti-modalposeestimation

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper sets out to show that a mobile robot can determine its global position inside a construction site from a single LiDAR scan, using nothing but a 3D mesh of the building and no real-world LiDAR training data. The authors train a diffusion model on synthetic scans generated by ray casting a simulated LiDAR into the mesh, so the model learns a distribution of plausible positions instead of a single estimate. That distribution is what lets the method survive perceptual aliasing—repeated rooms and symmetrical layouts that make individual scans ambiguous. Across five real floors in three buildings, the method reports 77% accuracy within 2 m on average and roughly half the mean error of the compared baselines. If correct, the result would give construction robots a practical way to re-localize after being dropped into an unknown pose, without markers or a prior guess.

What carries the argument

The load-bearing object is a conditional denoising diffusion probabilistic model used as a pose proposer. A PointNet++ point cloud encoder converts a LiDAR scan into a global feature vector; a transformer-based denoiser is trained to reverse Gaussian noise added to ground-truth positions, conditioned on that feature vector and a diffusion time step. During deployment the trained denoiser is sampled from many random starting positions to produce a set of pose hypotheses, which is the mechanism that captures multi-modal ambiguity. A second mechanism, fast global registration (FGR), renders a synthetic scan at each candidate, aligns it to the real scan, and uses the alignment fitness to pick and refine the final pose.

What would settle it

In a building with a mesh already available, keep one area geometrically unchanged and modify a neighboring area by adding a partition wall or dense clutter without updating the mesh; if the method's accuracy within 2 m stays high in the unchanged area but collapses in the modified area, the mesh-fidelity assumption is confirmed as the load-bearing premise. A cleaner version: measure acc@2m on scans collected before and after a renovation in the same building, using the same pre-renovation mesh for training; the after-renovation accuracy should fall unless the model has learned features that ignore the changed geometry.

Watch

Extended reading notes

Core claim

The central claim is that a diffusion model trained only on synthetic LiDAR scans can predict the global position of a real scan inside a building and can do so while honestly representing ambiguous alternatives. The model encodes the observed point cloud with a PointNet++ feature extractor and then runs a denoising diffusion process conditioned on that feature vector: starting from random positions, it iteratively removes learned noise to produce candidate positions. Because the process is stochastic, one scan yields many candidates, and the spread of those candidates reflects the localizability of the place. A fast global registration step then aligns a simulated scan at each candidate to the real scan and selects the best-fitting one. The evidence is the reported accuracy across five real-world floors: 77% of predictions land within 2 m, with 4-m accuracy above 88% on the two least aliased floors, and an oracle that always picks the best of the 25 candidates reaches 86-96% at 2 m, showing that the remaining error is largely in candidate selection, not in whether the correct position was proposed.

Load-bearing premise

The entire method presumes that the mesh used to synthesize training scans is an accurate, current representation of the building at deployment time; if walls or clutter have changed since the mesh was captured, the simulated scans no longer match the real scan and the learned position distribution can miss the true location.

Editorial extensions

If this is right

On a construction site with an up-to-date mesh, a robot can be re-localized from one LiDAR scan without any prior pose, GPS, or visual markers.
Ambiguous environments stop being fatal: instead of averaging symmetric or repeated rooms into one wrong answer, the model emits multiple hypotheses and a registration check picks the surviving mode.
The learned proposal distribution is far cheaper than exhaustive search; roughly 200 diffusion samples match the accuracy of more than 5000 grid-registered positions.
The variance of sampled positions doubles as a localizability map of the environment, flagging regions where a robot will struggle before it goes there.
Multi-floor operation is possible, though accuracy drops when similar hallways or rooms appear on different floors, since the model can assign a scan to the wrong floor.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The same synthetic-training recipe could be pointed at a BIM or CAD model rather than a laser-scanned mesh, making the approach usable before physical construction starts; the requirement is only a geometric surface to ray cast against.
The paper's own sensitivity caveat suggests a natural extension: adding simulated clutter, scaffolding, or movable partitions to the training mesh could make the learned features robust to the gap between mesh capture and deployment.
The multi-modal candidate distribution could be reused as a proposal distribution for a particle filter or as a signal for active localization, telling the robot where to move next to disambiguate symmetric rooms.
A direct test of the variance-to-error correlation is within reach: if per-scan variance of the 25 candidates predicts registration failure, then the localizability map could be converted into a per-scan confidence score.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Colleague,

Quick read: this is a competent applied robotics paper. The genuinely new bit is using a diffusion model, trained entirely on synthetic LiDAR ray-cast from a building mesh, to produce multi-modal position hypotheses for global relocalization in construction sites, then picking a hypothesis with fast global registration. It works about as well as claimed on five real floors, and the ablations actually support the design choices: diffusion beats regression at handling symmetric rooms, and N=25 with FGR beats N=1. The localizability map is a nice bonus.

That said, the paper has three real soft spots. First, Table I contains an impossible mean/median pair for the Hilti Scan Context Trajectory row (mean 0.14 m, median 9.75 m). That has to be a data-entry error; as printed it makes no sense. Second, the gap between FGR and the Oracle selector (e.g., Aesch 3 mean 5.20 m vs 1.51 m) is large, and the paper treats FGR as 'practical' without much discussion of how much performance is lost to a weak selection criterion. That is a limitation of the deployed pipeline and should be stated more prominently. Third, and most structurally: every synthetic training scan and FGR verification scan comes from the same mesh used at deployment time. The conclusion admits sensitivity to environment change, but no experiment measures degradation under mesh drift, added clutter, or wall changes. So the 77% acc@2m is scoped to 'same mesh, same time', which is a narrower claim than the intro's construction-site deployment framing. The stress-test note is on the mark here.

Minor stuff: no error bars or repeated runs, hyperparameters tuned on Hilti validation, no cross-building generalization test, and no code or data release. None of these are fatal, but they cap confidence.

Is it a real result? Yes. Sim-to-real transfer from mesh ray-casting to real scans is credible, and the comparison to grid and Scan Context baselines is fair. The paper deserves a serious referee and likely acceptance after fixing the table, discussing FGR loss, and explicitly scoping the mesh-fidelity assumption. I'd bring it to a reading group focused on robotic mapping or sim-to-real localization.

Referee Report

4 major / 4 minor

Summary. The paper proposes a diffusion-based LiDAR place recognition method for global re-localization on construction sites. A PointNet++ feature extractor encodes a LiDAR scan, and a transformer-based denoising diffusion model is trained on synthetic scans generated by ray casting in a real building mesh to predict the conditional distribution of the scan's global position. During deployment, multiple candidate positions are sampled from the learned distribution, and fast global registration (FGR) selects and refines the best candidate. The method is evaluated on five real-world floors across three buildings and is reported to achieve 77% accuracy within 2 m on average, outperforming regression, grid search, and Scan Context baselines, with ablations on the number of candidates, selection criterion, and sampling strategy. The paper also analyzes multi-modal predictions under perceptual aliasing and proposes a localizability map from prediction variance.

Significance. If the results hold, the paper makes a useful contribution: it demonstrates that a diffusion model trained entirely on synthetic LiDAR data can produce multi-modal position estimates in perceptually aliased indoor environments, and that the resulting candidate set is more informative than regression or dense grid sampling. The experimental breadth (five floors, three buildings, multiple baselines, and ablations) is a strength, and the authors explicitly discuss the limitations of their approach. The finding that diffusion candidates are more efficient than grid candidates (Fig. 7) is practically relevant. However, the significance is tempered by the reliance on an accurate and current mesh for both training and verification, by an inconsistent baseline entry in Table I, and by ambiguity about the handling of yaw/heading in the training distribution. These issues need to be resolved before the claims can be taken at face value.

major comments (4)

[Table I (Hilti, Scan Context Trajectory row)] The reported mean error of 0.14 m and median error of 9.75 m cannot both be correct: if at least half of the errors are at least 9.75 m, the mean error must be at least 4.875 m. The acc@4m=0.15 and acc@2m=0.04 values also contradict a mean of 0.14 m. This row is used in the claim that the proposed method outperforms Scan Context even with privileged trajectory information, so the numbers must be corrected or the row must be relabeled.
[Table III and Section IV-E (Selection Criterion)] The FGR selection step is a substantial bottleneck: replacing FGR with an Oracle improves mean error from 3.69 m to 0.96 m on LEE H and from 5.20 m to 1.51 m on Aesch 3, with acc@2m rising by more than 20 percentage points in several floors. Since FGR is an integral part of the proposed pipeline, the paper should investigate why FGR so often selects the wrong mode, or replace it with a more robust verification step; without this, the method's practical robustness is weaker than the multi-modal prediction capability suggests.
[Section III-C (Implementation Details)] The training description only specifies rotations 'around axes perpendicular to the ground plane surface normal' in the range ±20 degrees, which covers pitch and roll tilts but not yaw/heading. Since the network operates on raw point-cloud features that are not yaw-invariant, and the predicted quantity is (x,y,z), the paper must clarify whether yaw is sampled uniformly, whether scans are canonicalized to a known heading, or whether the system assumes a known orientation. This detail is load-bearing for the general 'global re-positioning' claim.
[Section V (Conclusion)] The conclusion acknowledges that the approach is sensitive to changes in the environment between mesh acquisition and deployment, but no experiment measures this sensitivity. Both the synthetic training scans and the FGR verification templates are generated from the same mesh, so all results are contingent on the mesh being exactly current. A temporal split or a mesh-perturbation experiment (for example, adding or removing walls, doors, or clutter) is needed to support the 'robust' claim in the title and to define the practical deployment envelope.

minor comments (4)

[Section III-A (Regression with Diffusion)] The notation uses α_t for both 1−β_t and the cumulative product ∏_{s=1}^t α_s, which is likely to confuse readers; introduce \bar{α}_t for the cumulative product, as in the standard DDPM notation.
[Fig. 2 and Section III-B] The label 'FGR argmax' in the pipeline figure is ambiguous; clarify that FGR selects the candidate with the highest registration fitness, not an argmax over positions.
[Section IV-B] The paper does not compare against the closely related diffusion-based localization method DiffLoc [9]; a comparison or an explicit discussion of the differences in problem setup and sensor representation would strengthen the positioning of the contribution.
[Section V] There is a typo in the conclusion: 'occured' should be 'occurred'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the trained diffusion model is evaluated on real LiDAR scans against ground truth, and the synthetic training data from the mesh is an input assumption rather than a relabeled prediction.

full rationale

The derivation chain is self-contained. The model is trained to denoise positions conditioned on features extracted from synthetic LiDAR scans that are ray-cast at random poses in the mesh. The central quantity predicted, p(x|c_x), is learned with a standard DDPM objective; no fitted constant is later relabeled as a predicted position. Evaluation uses real LiDAR scans from five floors with ground-truth trajectories, so the reported accuracy is not determined by construction. The FGR verification step also uses point clouds simulated in the same mesh, but this is the intended map-based reference frame for global re-localization and is not an output of the diffusion model; the real scan is matched against mesh-rendered candidates and the final pose is measured against externally collected ground truth. The acknowledged sensitivity to changes between mesh acquisition and deployment is a scope limitation, not a circularity: it restricts external validity without making the result equal to the training input. The only mild selection step is tuning feature-extractor hyperparameters on the Hilti validation set, which weakens the independence of the Hilti row but does not make any result true by definition. No self-citation chain, uniqueness theorem, or ansatz is used in a load-bearing way; the cited baselines and diffusion framework are external and standard.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

Free parameters are mostly standard hyperparameters plus the candidate count N. The central axioms are domain assumptions about mesh fidelity and sim-to-real transfer, plus reliance on FGR as a selector. No new physical entities, forces, or conserved quantities are introduced.

free parameters (4)

Number of sampled candidates N = 25
Chosen by authors; accuracy improves with more candidates (Table III, Fig. 7), so reported results are tied to N=25.
Diffusion time steps T = 100
Hyperparameter of the DDPM with linear beta schedule from 0.0001 to 0.02.
Point cloud subsample size = 4096
Selected to reduce compute; feature extraction and model behavior depend on it.
Training tilt range = -20 to 20 degrees
Rotations about horizontal axes are augmented within this range, a hand-set distribution choice that may be appropriate for ground robots.

assumptions (4)

domain assumption Ray-casting synthetic LiDAR scans in the mesh produces point clouds representative of real LiDAR scans in the same environment.
The entire training pipeline relies on this; pre-processing removes distant points to reduce sim-to-real gap, and the conclusion admits sensitivity to environment changes not present in the mesh.
domain assumption Fast global registration's fitness score can select the correct candidate often enough for the final pose estimate.
Used in deployment and in reported metrics; Table III's Oracle comparison shows FGR is clearly worse than an oracle selector, so this assumption is only partially satisfied.
standard math DDPM denoising and the noise schedule from Ho et al. accurately approximate the posterior position distribution.
The method imports standard diffusion results (Eq. 1-3), not proved in this paper.
domain assumption The building mesh is globally consistent and in the reference frame used for evaluation.
Training and evaluation positions are defined relative to the mesh; inconsistent or misaligned meshes would invalidate localization.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Diffusion Based Robust LiDAR Place Recognition." pith.science (2026). https://pith.science/paper/3VMZXS3K

@misc{pith2026250412412,
  author       = {Pith},
  title        = {Pith review of: Diffusion Based Robust LiDAR Place Recognition},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3VMZXS3K}},
  note         = {Machine review of arXiv:2504.12412}
}

read the original abstract

Mobile robots on construction sites require accurate pose estimation to perform autonomous surveying and inspection missions. Localization in construction sites is a particularly challenging problem due to the presence of repetitive features such as flat plastered walls and perceptual aliasing due to apartments with similar layouts inter and intra floors. In this paper, we focus on the global re-positioning of a robot with respect to an accurate scanned mesh of the building solely using LiDAR data. In our approach, a neural network is trained on synthetic LiDAR point clouds generated by simulating a LiDAR in an accurate real-life large-scale mesh. We train a diffusion model with a PointNet++ backbone, which allows us to model multiple position candidates from a single LiDAR point cloud. The resulting model can successfully predict the global position of LiDAR in confined and complex sites despite the adverse effects of perceptual aliasing. The learned distribution of potential global positions can provide multi-modal position distribution. We evaluate our approach across five real-world datasets and show the place recognition accuracy of 77% +/-2m on average while outperforming baselines at a factor of 2 in mean error.

Figures

Figures reproduced from arXiv: 2504.12412 by the authors.

**Figure 2.** Overview of the proposed place recognition module. During training, we sample random poses inside the mesh and ray cast the LiDAR pattern [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗

**Figure 3.** The datasets used in the evaluation are shown. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: Trajectories of the different datasets are shown. The colors correspond to the error between the predicted and the ground truth positions. The grey [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]

**Figure 5.** Figure 5: Distribution of candidate points. The best guess is found by [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]

**Figure 8.** Figure 8: Localizability map of Aesch Floor 3 is provided. The color bar [PITH_FULL_IMAGE:figures/full_fig_p006_8.png]

**Figure 7.** Figure 7: Error distributions for different numbers of potential positions. [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 19 canonical work pages

[1]

NetVLAD: CNN architecture for weakly supervised place recognition

R. Arandjelovic, P. Gron ´at, A. Torii, T. Pajdla, and J. Sivic. Netvlad: CNN architecture for weakly supervised place recognition. CoRR, abs/1511.07247, 2015

work page Pith review arXiv 2015
[2]

Bosse and R

M. Bosse and R. Zlot. Place recognition using keypoint voting in large 3d lidar datasets. In 2013 IEEE International Conference on Robotics and Automation, pages 2677–2684, 2013

work page 2013
[3]

Del Duca and C

G. Del Duca and C. Machado. Assessing the quality of the leica blk2go mobile laser scanner versus the focus 3d s120 static terrestrial laser scanner for a preliminary study of garden digital surveying. Heritage, 6(2):1007–1027, 2023

work page 2023
[4]

simultaneous localisation and map- ping (slam): Part i the essential algorithms

H. Durrant-Whyte and T. Bailey. “simultaneous localisation and map- ping (slam): Part i the essential algorithms”. Robotics & Automation Magazine, 13, 01 2006

work page 2006
[5]

J. Ho, A. Jain, and P. Abbeel. Denoising diffusion probabilistic models, 2020

2020
[6]

Ho and T

J. Ho and T. Salimans. Classifier-free diffusion guidance. CoRR, abs/2207.12598, 2022

arXiv 2022
[7]

Kim and A

G. Kim and A. Kim. Scan context: Egocentric spatial descriptor for place recognition within 3d point cloud map. In 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pages 4802–4809, 2018

work page 2018
[8]

MinkLoc3D: Point Cloud Based Large-Scale Place Recognition

J. Komorowski. Minkloc3d: Point cloud based large-scale place recognition. CoRR, abs/2011.04530, 2020

work page Pith review arXiv 2011

Show all 25 references

[9]

W. Li, Y . Yang, S. Yu, G. Hu, C. Wen, M. Cheng, and C. Wang. Diffloc: Diffusion model for outdoor lidar localization. In Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 15045–15054, 2024

2024
[10]

L. Luo, S. Zheng, Y . Li, Y . Fan, B. Yu, S. Cao, and H. Shen. Bevplace: Learning lidar-based place recognition using bird’s eye view images, 2023

2023
[11]

Luo and W

S. Luo and W. Hu. Diffusion probabilistic models for 3d point cloud generation. CoRR, abs/2103.01458, 2021

2021 arXiv
[12]

C.R. Qi, H. Su, K. Mo, and L.J. Guibas. Pointnet: Deep learning on point sets for 3d classification and segmentation, 2017

2017
[13]

C.R. Qi, L. Yi, H. Su, and L.J. Guibas. Pointnet++: Deep hier- archical feature learning on point sets in a metric space. CoRR, abs/1706.02413, 2017

2017 arXiv
[14]

Z. Qiao, H. Huang, C. Liu, S. Shen, F. Zhang, and H. Yin. Speak the same language: Global lidar registration on bim using pose hough transform, 2024

2024
[15]

Rombach, A

R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer. High-resolution image synthesis with latent diffusion models. CoRR, abs/2112.10752, 2021

2021 arXiv
[16]

Saharia, W

C. Saharia, W. Chan, S. Saxena, L. Li, J. Whang, E. Denton, S.K.S. Ghasemipour, B.K. Ayan, S.S. Mahdavi, R.G. Lopes, T. Salimans, J. Ho, D.J. Fleet, and M. Norouzi. Photorealistic text-to-image diffusion models with deep language understanding, 2022

2022
[17]

Siegwart, I.R

R. Siegwart, I.R. Nourbakhsh, and D. Scaramuzza. Introduction to Autonomous Mobile Robots . The MIT Press, 2nd edition, 2011

2011
[18]

J. Song, C. Meng, and S. Ermon. Denoising diffusion implicit models. CoRR, abs/2010.02502, 2020

2010 arXiv
[19]

Uy and G.H

M.A. Uy and G.H. Lee. Pointnetvlad: Deep point cloud based retrieval for large-scale place recognition. CoRR, abs/1804.03492, 2018

2018 arXiv
[20]

W. Wang, B. Wang, P. Zhao, C. Chen, R. Clark, B. Yang, A. Markham, and N. Trigoni. Pointloc: Deep pose regressor for lidar point cloud localization, 2021

2021
[21]

Xu and B

X. Xu and B. Garcia de Soto. On-site autonomous construction robots: A review of research areas, technologies, and suggestions for advancement. In Proceedings of the 37th International Symposium on Automation and Robotics in Construction (ISARC) , pages 385– 392, Kitakyushu, J...

2020
[22]

L. Yang, Z. Zhang, Y . Song, S. Hong, R. Xu, Y . Zhao, W. Zhang, B. Cui, and M.H. Yang. Diffusion models: A comprehensive survey of methods and applications, 2024

2024
[23]

Zhang, M

L. Zhang, M. Helmberger, L.F.T. Fu, D. Wisth, M. Camurri, D. Scara- muzza, and M. Fallon. Hilti-oxford dataset: A millimeter-accurate benchmark for simultaneous localization and mapping. IEEE Robotics and Automation Letters , 8(1):408–415, 2023

2023
[24]

Zhang, P

Y . Zhang, P. Shi, and J. Li. Lidar-based place recognition for autonomous driving: A survey, 2023

2023
[25]

Q.Y . Zhou, J. Park, and V . Koltun. Fast global registration. In B. Leibe, J. Matas, N. Sebe, and M. Welling, editors, Computer Vision – ECCV 2016, pages 766–782, Cham, 2016. Springer International Publishing

2016

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.

[1] [1]

NetVLAD: CNN architecture for weakly supervised place recognition

R. Arandjelovic, P. Gron ´at, A. Torii, T. Pajdla, and J. Sivic. Netvlad: CNN architecture for weakly supervised place recognition. CoRR, abs/1511.07247, 2015

work page Pith review arXiv 2015

[2] [2]

Bosse and R

M. Bosse and R. Zlot. Place recognition using keypoint voting in large 3d lidar datasets. In 2013 IEEE International Conference on Robotics and Automation, pages 2677–2684, 2013

work page 2013

[3] [3]

Del Duca and C

G. Del Duca and C. Machado. Assessing the quality of the leica blk2go mobile laser scanner versus the focus 3d s120 static terrestrial laser scanner for a preliminary study of garden digital surveying. Heritage, 6(2):1007–1027, 2023

work page 2023

[4] [4]

simultaneous localisation and map- ping (slam): Part i the essential algorithms

H. Durrant-Whyte and T. Bailey. “simultaneous localisation and map- ping (slam): Part i the essential algorithms”. Robotics & Automation Magazine, 13, 01 2006

work page 2006

[5] [5]

J. Ho, A. Jain, and P. Abbeel. Denoising diffusion probabilistic models, 2020

2020

[6] [6]

Ho and T

J. Ho and T. Salimans. Classifier-free diffusion guidance. CoRR, abs/2207.12598, 2022

arXiv 2022

[7] [7]

Kim and A

G. Kim and A. Kim. Scan context: Egocentric spatial descriptor for place recognition within 3d point cloud map. In 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pages 4802–4809, 2018

work page 2018

[8] [8]

MinkLoc3D: Point Cloud Based Large-Scale Place Recognition

J. Komorowski. Minkloc3d: Point cloud based large-scale place recognition. CoRR, abs/2011.04530, 2020

work page Pith review arXiv 2011