Pith. sign in

REVIEW 4 major objections 5 minor 92 references

Advancing ALS Applications with Large-Scale Pre-training: Dataset Development and Downstream Assessment

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read A 184-billion-point pre-training dataset makes ALS models transfer to downstream tasks.

desk verdict Useful dataset and honest limitations, but the transferability claims are undercut by a missing overlap check and tiny DALES gains. read the letter →

arxiv 2501.05095 v1 pith:FSB2JQCF submitted 2025-01-09 cs.CV cs.AI

Haoyi Xiu , Xin Liu , Taehoon Kim , Kyoung-Sook Kim This is my paper · ORCID
classification cs.CVcs.AI
keywords airbornelaserscanningpre-training3Dpointcloudsmaskedautoencodersgeospatialsamplinglandcoverdigitalelevationmodelfoundationmodels
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

The paper tries to establish that the pre-train-then-fine-tune paradigm, which has already transformed satellite remote sensing, can work for airborne laser scanning (ALS) point clouds. To show this, the authors build a 184-billion-point dataset from a national LiDAR program, using a geospatial sampling method that balances land cover and terrain slope instead of picking tiles randomly. They then pre-train a masked autoencoder called BEV-MAE on this dataset and fine-tune it for tree species classification, terrain scene recognition, and urban semantic segmentation. Across all three tasks, the pre-trained models beat models trained from scratch, and in segmentation performance rises with dataset scale only when their sampling method is used. If true, this gives the ALS community a reusable pre-training resource and a recipe for building diverse LiDAR datasets from public archives.

What carries the argument

The central mechanism is a geospatial sampling procedure: for each 3DEP LiDAR project, the land-cover map and DEM-derived slope classification are cropped to the project boundary, divided into 500 m by 500 m patches, labeled by the majority land cover and slope class, and selected through inverse-probability sampling over the joint land-cover-slope distribution, restricted to the Developed and Forest classes. This yields 73,762 tiles, which are used to pre-train BEV-MAE, a masked autoencoder for outdoor 3D point clouds that masks bird's-eye-view pillars and reconstructs point coordinates and density through a sparse 3D convolutional encoder-decoder. The sampling method is what the paper credits for the consistent scaling gains, distinguishing it from random sampling at matched scale.

What would settle it

Remove any 3DEP pre-training tiles that intersect the DALES test area in Dayton, Ohio, or the OpenGF-derived terrain-scene tiles, retrain the pre-trained model on the remaining tiles, and re-run the fine-tuning; if the performance gap over scratch collapses or shrinks to noise, the central transferability claim fails.

Watch

Extended reading notes

Core claim

The paper claims that a large-scale ALS point-cloud dataset built from the 3DEP source, sampled to balance land cover (Developed vs. Forest) and terrain slope (Flat/Sloped/Steep), supports effective self-supervised pre-training with BEV-MAE. The pre-trained models outperform their scratch counterparts on tree species classification (mIoU 75.6 vs. 72.2), terrain scene recognition (87.4 vs. 86.6), and urban semantic segmentation (78.2 vs. 77.9 at 40 samples per project). It further claims that scaling the dataset using the geospatial sampling method consistently improves segmentation performance, while scaling with random sampling does not, and that pre-training on the smaller OpenGF dataset underperforms. These results are offered as evidence that both dataset scale and a diversity-aware sampling strategy are essential for successful ALS pre-training.

Load-bearing premise

The reported gains come from transferable representations rather than from spatial overlap between the pre-training tiles and the downstream evaluation areas.

Editorial extensions

If this is right

  • Downstream ALS tasks with scarce labels can start from the released pre-trained weights instead of training from scratch, potentially reducing annotation costs.
  • Large public LiDAR archives become usable for pre-training only when paired with a diversity-aware sampling strategy; random sampling at similar scale does not yield the same scaling gains.
  • Terrain scene recognition, a task with few public datasets, gains a new benchmark derived from OpenGF and a reported pre-training advantage.
  • The 184-billion-point, 73,762-tile dataset becomes a scale reference point for future ALS foundation-model research.

Reading between the lines

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

  • My inference: the same land-cover and slope sampling recipe should transfer to other national LiDAR archives, such as the Dutch AHN data, and success there would support the claim that class-balanced diversity rather than archive specificity drives the gains.
  • My inference: the especially large improvement on Black locust and Douglas fir in the tree-species task may partly reflect species-range overlap with the pre-training geography, so a cleaner test would restrict evaluation to sites outside the pre-training footprint.
  • My inference: the modest segmentation gain points to a concrete next experiment: pre-training with a reconstruction loss that preserves fine geometry, such as an added perceptual or point-normal loss, should increase the segmentation benefit if the paper's stated hypothesis is correct.
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.

Referee Report

4 major / 5 minor

Summary. The paper constructs a large-scale ALS point-cloud pre-training dataset from USGS 3DEP data covering the conterminous United States, using a proposed geospatial sampling method based on land cover (NLCD) and slope (DEM) maps. It adopts BEV-MAE as a self-supervised pre-training model and evaluates fine-tuning on three downstream tasks: tree species classification on PureForest, terrain scene recognition on a newly created OpenGF-derived dataset, and urban semantic segmentation on DALES. The central claims are that the pre-trained models outperform scratch counterparts across all downstream tasks and that the geospatial sampling method yields consistent improvements as dataset size grows, whereas random sampling does not.

Significance. If the claims hold, the dataset and pre-trained models would be a useful community resource for a relatively underexplored area: large-scale self-supervised pre-training for ALS point clouds rather than satellite imagery. The paper contributes a substantial dataset (73,762 tiles, about 184 billion points), a reproducible sampling methodology, and an open-source code/model release. The evaluation across three different downstream tasks is a strength, and the PureForest tree-species experiment is particularly valuable because it is geographically disjoint from the CONUS pre-training data, supporting transferability. However, the strength of the empirical evidence is currently uneven: the DALES segmentation gains are small and lack statistical validation, the sampling-strategy comparison is limited to a single task, and no decontamination analysis is reported between pre-training tiles and downstream evaluation areas. These gaps weaken the generality of the abstract's claims until addressed.

major comments (4)
  1. [Sections 3.1, 5.1.2, 5.1.3] The manuscript never reports a decontamination check between the pre-training tiles and the downstream evaluation tiles. Section 3.1 states that the pre-training data are 3DEP LiDAR point clouds covering the conterminous United States; Section 5.1.3 evaluates on DALES, which is collected in Dayton, Ohio, and Section 5.1.2 builds terrain scenes from OpenGF, an international collection that may include U.S. sites. No exclusion of pre-training tiles overlapping the DALES or OpenGF evaluation areas is described, and the coordinates of OpenGF scenes are not given. Because the pre-training task is BEV occupancy reconstruction, exact or nearby duplicate tiles could inflate downstream mIoU through geometry memorization rather than transferable representation. The authors should report the minimum distance between every pre-training tile and every evaluation tile, remove any overlapping or near-duplicate tiles, and rerun the fine-tuning comparisons.
  2. [Section 6.2.3, Table 13] The DALES segmentation results do not support the word "significantly" in the abstract. Table 13 reports a single mIoU per setting (no number of runs is given, in contrast to Tables 9 and 11, which explicitly report averages of three runs), and the differences among scratch (77.9), random sampling variants (77.6-77.8), and the proposed sampling variants (77.7-78.2) are 0.1-0.3 mIoU, which is within typical run-to-run noise for semantic segmentation. The text should report the number of seeds, the mean and standard deviation, and a paired significance test, or explicitly refrain from claiming statistical significance. Without this, the claim that "the pre-trained models significantly outperform their scratch counterparts across all downstream tasks" is not established for DALES.
  3. [Abstract, Section 6.2.3] The scaling claim is over-stated relative to the evidence. The "consistently enhances" conclusion rests on three dataset sizes (10, 20, and 40 samples per project) on a single downstream task (DALES), and the differences are 77.7 -> 78.0 -> 78.2 mIoU with no uncertainty intervals. Moreover, at 10 samples per project the proposed-sampling pre-trained model (77.7) is below the scratch model (77.9), which contradicts the abstract's unconditional statement that pre-trained models outperform scratch across all downstream tasks. The authors should qualify the claim to the specific settings and add error bars or additional scaling points.
  4. [Section 3.2.3, Table 3, Section 6.2.3] The sampling-strategy evaluation is narrower than the paper's conclusions. The geospatial sampling method is compared against random sampling only for DALES segmentation in Table 13; tree species classification and terrain scene recognition use only the full 40-sample-per-project pre-trained model. Furthermore, the pre-training dataset deliberately includes only "Developed" and "Forest" land cover classes, so statements about capturing "diverse land cover" (Abstract, Section 3.2) should be qualified. If the claim is that the geospatial sampling method is broadly effective, it needs either more downstream tasks with the random baseline or a clear statement that the method is evaluated for developed-area segmentation only.
minor comments (5)
  1. [Section 4 and Section 6.2.1] There is an inconsistency in the voxel size: Section 4 and Section 5.1.1 state a pre-training/fine-tuning voxel size of 0.6 m, but the discussion after Table 10 says "we used 0.06 m which results in much higher resolution" relative to the 0.25 m baseline. The authors should correct this; if the voxel size is actually 0.6 m, the comparison with the 0.25 m baseline is not a higher-resolution comparison.
  2. [Tables 9 and 10] Table 9 reports mIoU as an average of three runs, while Table 10 reports the "Mean" row for the best run; the two numbers differ (75.6 vs. 77.6 for the proposed model), which is confusing without an explicit explanation of which value is the headline metric. Please add a sentence clarifying the relationship between the two tables.
  3. [Section 5.1.2] The terrain scene recognition test set is quite small: 675 test tiles are derived from only 27 distinct 500 m x 500 m scenes, and the nine classes include classes with very few samples. A short discussion of the small-sample caveat and per-class confidence intervals would strengthen the interpretation of Table 12.
  4. [Section 3.2.3, Table 3] The manuscript should state explicitly that the pre-training dataset includes only two Level I land cover classes (Developed and Forest) and explain how this choice affects the generality of the dataset for other downstream tasks such as water, barren, or wetland applications.
  5. [Throughout] There are several typos and grammatical errors that should be fixed, including "develope" (Section 7), "Bold test shows" (captions of Tables 9-13), "an memory efficient variant" (Section 4), "We mainly focuses" (Section 2.3.1), and "align roughly well" (Section 6.1).

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: pre-training and downstream evaluation are independent.

full rationale

The paper's derivation chain is not circular. The pre-training dataset is constructed by geospatial sampling of USGS 3DEP LiDAR tiles using NLCD land cover and DEM slope maps (Sections 3.2-3.3), and BEV-MAE is pre-trained self-supervised on unlabeled coordinates with a masked-autoencoding objective (Section 4). Downstream evaluations use external labeled benchmarks: PureForest for tree species classification, a terrain-scene split derived from OpenGF, and DALES for semantic segmentation. No fitted parameter is renamed as a prediction: the only fitted quantities are the BEV-MAE weights, and the downstream metrics (mIoU, OA, Eqs. 1-3) are computed on independently labeled test folds. The reported gains (e.g., Table 9: 75.6 vs 72.2 mIoU; Table 11: 87.4 vs 86.6; Table 13: 78.2 vs 77.9) are empirical outcomes, not definitions. The OpenGF-pretrained baseline in Table 13 is evaluated on DALES, not on the OpenGF-derived terrain-scene test set, so there is no construction-level equivalence between pretraining source and evaluation set. The paper's post-hoc explanation for tree-species gains (U.S.-native species in pretraining data) is speculative but not circular, and no load-bearing claim rests on a self-citation chain; mentions of the authors' own prior work (e.g., reference [4]) are not used to justify the central transferability claim. The potential spatial overlap between 3DEP CONUS tiles and DALES/OpenGF test areas is a data-contamination/correctness risk, not a definitional or self-citational circularity, and therefore does not raise the circularity score.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The central claim rests on design choices in the sampling method (only two land cover classes, slope thresholds, per-project sample caps, voxel resolution) and on assumptions about data quality, regional transferability, and evaluation decontamination. No new physical entities are introduced.

free parameters (5)
  • Number of sampled tiles per LiDAR project = 40 (variants 10, 20 tested)
    Chosen cap on data collection; the scaling experiments vary this number and the central sampling-advantage claim depends on it.
  • Slope class thresholds = 5 degrees and 17 degrees
    USDA-adapted thresholds partition terrain into Flat/Sloped/Steep and define the pre-training distribution.
  • Land cover classes selected = Developed and Forest only
    Only two Level I NLCD classes are sampled; other classes are excluded, which limits the land-cover diversity claim.
  • Voxel resolution during pre-training = 0.6 m (stated as 0.06 m in Section 6.2.1)
    Spatial resolution affects learned features; the paper is internally inconsistent about this value.
  • Maximum voxels per tile = 200,000
    Computational cap for pre-training and fine-tuning; affects which points are seen.
assumptions (5)
  • domain assumption NLCD land cover labels are sufficiently accurate for tile selection
    Sampling relies on Level I classes merged from NLCD2021; the paper cites the NLCD2019 accuracy assessment and uses only the high-accuracy Developed and Forest classes (Section 3.2.1).
  • domain assumption 3DEP LiDAR data are consistent enough to be treated as one unified dataset
    The paper relies on 3DEP base specifications for consistent acquisition and treats the corpus as a single dataset (Section 3.1, citing [11]).
  • domain assumption Pre-training on CONUS 3DEP data transfers to evaluation scenes in France and other regions
    PureForest tree species data are from southern France; transferability across regions and sensors is assumed without dedicated analysis (Sections 5.1.1 and 6.2.1).
  • domain assumption BEV-MAE is an appropriate SSL baseline for ALS data
    The paper adopts BEV-MAE without architectural modifications and assumes its outdoor LiDAR design suits ALS point clouds (Section 4).
  • domain assumption No overlap exists between pre-training tiles and downstream evaluation areas
    No decontamination is described; 3DEP covers CONUS and DALES/OpenGF scenes may coincide with sampled tiles (Sections 3.3 and 6.2.3).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Advancing ALS Applications with Large-Scale Pre-training: Dataset Development and Downstream Assessment." pith.science (2026). https://pith.science/paper/FSB2JQCF

@misc{pith2026250105095,
  author       = {Pith},
  title        = {Pith review of: Advancing ALS Applications with Large-Scale Pre-training: Dataset Development and Downstream Assessment},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FSB2JQCF}},
  note         = {Machine review of arXiv:2501.05095}
}
read the original abstract

The pre-training and fine-tuning paradigm has revolutionized satellite remote sensing applications. However, this approach remains largely underexplored for airborne laser scanning (ALS), an important technology for applications such as forest management and urban planning. In this study, we address this gap by constructing a large-scale ALS point cloud dataset and evaluating its impact on downstream applications. Our dataset comprises ALS point clouds collected across the contiguous United States, provided by the United States Geological Survey's 3D Elevation Program. To ensure efficient data collection while capturing diverse land cover and terrain types, we introduce a geospatial sampling method that selects point cloud tiles based on land cover maps and digital elevation models. As a baseline self-supervised learning model, we adopt BEV-MAE, a state-of-the-art masked autoencoder for 3D outdoor point clouds, and pre-train it on the constructed dataset. The pre-trained models are subsequently fine-tuned for downstream tasks, including tree species classification, terrain scene recognition, and point cloud semantic segmentation. Our results show that the pre-trained models significantly outperform their scratch counterparts across all downstream tasks, demonstrating the transferability of the representations learned from the proposed dataset. Furthermore, we observe that scaling the dataset using our geospatial sampling method consistently enhances performance, whereas pre-training on datasets constructed with random sampling fails to achieve similar improvements. These findings highlight the utility of the constructed dataset and the effectiveness of our sampling strategy in the pre-training and fine-tuning paradigm. The source code and pre-trained models will be made publicly available at \url{https://github.com/martianxiu/ALS_pretraining}.

Figures

Figures reproduced from arXiv: 2501.05095 by the authors.

Figure 1
Figure 1. Overview of the dataset development procedure: Land cover data, DEM, and [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. LiDAR point cloud boundaries used in this study are shown with randomly assigned colors for the boundary polygons. The boundary data were downloaded from [71] on June 27, 2024. 3.2. Geospatial sampling While 3DEP offers abundant resources for conducting pre-training, utiliz￾ing the entire dataset (>300TB) is practically infeasible. Therefore, a sam￾pling strategy is required to extract representative data from U.S. … view at source ↗
Figure 3
Figure 3. The upper left figure displays the land cover map derived from the Anderson [PITH_FULL_IMAGE:figures/full_fig_p013_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Random samples of the dataset. Top: point cloud tiles labeled as “Developed”. [PITH_FULL_IMAGE:figures/full_fig_p018_4.png]
Figure 5
Figure 5. Figure 5: Overview of the pre-training and fine-tuning using BEV-MAE. [PITH_FULL_IMAGE:figures/full_fig_p020_5.png]
Figure 6
Figure 6. Figure 6: Some examples from the terrain scene recognition dataset developed based on [PITH_FULL_IMAGE:figures/full_fig_p022_6.png]
Figure 7
Figure 7. Figure 7: Results of coordinate reconstruction. GT represents the Ground Truth, while [PITH_FULL_IMAGE:figures/full_fig_p025_7.png]
Figure 8
Figure 8. Figure 8: Results of density reconstruction [PITH_FULL_IMAGE:figures/full_fig_p026_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

92 extracted references · 64 canonical work pages

  1. [1]

    N. Qin, W. Tan, L. Ma, D. Zhang, J. Li, Opengf: An ultra-large- scale ground filtering dataset built upon open als point clouds around the world, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021, pp. 1082–1091. 31

    2021

  2. [2]

    Gaydon, F

    C. Gaydon, F. Roche, Pureforest: A large-scale aerial lidar and aerial imagery dataset for tree species classification in monospecific forests, arXiv preprint arXiv:2404.12064 (2024)

    arXiv 2024

  3. [3]

    Varney, V

    N. Varney, V. K. Asari, Q. Graehling, Dales: A large-scale aerial lidar data set for semantic segmentation, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops, 2020, pp. 186–187

    2020

  4. [4]

    H. Xiu, X. Liu, W. Wang, K.-S. Kim, T. Shinohara, Q. Chang, M. Mat- suoka, Ds-net: A dedicated approach for collapsed building detection from post-event airborne point clouds, International Journal of Applied Earth Observation and Geoinformation 116 (2023) 103150

    2023

  5. [5]

    Brown, B

    T. Brown, B. Mann, N. Ryder, M. Subbiah, J. D. Kaplan, P. Dhariwal, A. Neelakantan, P. Shyam, G. Sastry, A. Askell, et al., Language models are few-shot learners, Advances in neural information processing systems 33 (2020) 1877–1901

    2020

  6. [6]

    Radford, J

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clark, et al., Learning transfer- able visual models from natural language supervision, in: International conference on machine learning, PMLR, 2021, pp. 8748–8763

    2021

  7. [7]

    Bommasani, D

    R. Bommasani, D. A. Hudson, E. Adeli, R. Altman, S. Arora, S. von Arx, M. S. Bernstein, J. Bohg, A. Bosselut, E. Brunskill, et al., On the opportunities and risks of foundation models, arXiv preprint arXiv:2108.07258 (2021)

    arXiv 2021

  8. [8]

    X. Guo, J. Lao, B. Dang, Y. Zhang, L. Yu, L. Ru, L. Zhong, Z. Huang, K. Wu, D. Hu, et al., Skysense: A multi-modal remote sensing foun- dation model towards universal interpretation for earth observation im- agery, in: Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, 2024, pp. 27672–27683

    2024

Show all 92 references
  1. [9]

    Mendieta, B

    M. Mendieta, B. Han, X. Shi, Y. Zhu, C. Chen, Towards geospatial foundation models via continual pretraining, in: Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 16806–16816. 32

    2023

  2. [10]

    D. Hong, B. Zhang, X. Li, Y. Li, C. Li, J. Yao, N. Yokoya, H. Li, P. Ghamisi, X. Jia, et al., Spectralgpt: Spectral remote sensing foun- dation model, IEEE Transactions on Pattern Analysis and Machine Intelligence (2024)

    2024

  3. [11]

    Stoker, B

    J. Stoker, B. Miller, The accuracy and consistency of 3d elevation pro- gram data: A systematic analysis, Remote Sensing 14 (2022) 940

    2022

  4. [12]

    URL: https://www.ahn.nl/, accessed: 2024-12-26

    Actueel Hoogtebestand Nederland, Actueel hoogtebestand nederland (ahn), n.d. URL: https://www.ahn.nl/, accessed: 2024-12-26

    2024

  5. [13]

    Wickham, S

    J. Wickham, S. V. Stehman, D. G. Sorenson, L. Gass, J. A. Dewitz, Thematic accuracy assessment of the nlcd 2019 land cover for the conter- minous united states, GIScience & Remote Sensing 60 (2023) 2181143

    2023

  6. [14]

    Z. Lin, Y. Wang, S. Qi, N. Dong, M.-H. Yang, Bev-mae: Bird’s eye view masked autoencoders for point cloud pre-training in autonomous driving scenarios, in: Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, 2024, pp. 3531–3539

    2024

  7. [15]

    H. Liu, C. Li, Q. Wu, Y. J. Lee, Visual instruction tuning, Advances in neural information processing systems 36 (2024)

    2024

  8. [16]

    Ayush, B

    K. Ayush, B. Uzkent, C. Meng, K. Tanmay, M. Burke, D. Lobell, S. Er- mon, Geography-aware self-supervised learning, in: Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 10181–10190

    2021

  9. [17]

    Manas, A

    O. Manas, A. Lacoste, X. Gir´ o-i Nieto, D. Vazquez, P. Rodriguez, Sea- sonal contrast: Unsupervised pre-training from uncurated remote sens- ing data, in: Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 9414–9423

    2021

  10. [18]

    U. Mall, B. Hariharan, K. Bala, Change-aware sampling and contrastive learning for satellite images, in: Proceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition, 2023, pp. 5261–5270

    2023

  11. [19]

    X. Chen, H. Fan, R. Girshick, K. He, Improved baselines with momen- tum contrastive learning, arXiv preprint arXiv:2003.04297 (2020). 33

    2020 arXiv

  12. [20]

    K. He, X. Chen, S. Xie, Y. Li, P. Doll´ ar, R. Girshick, Masked autoen- coders are scalable vision learners, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 16000– 16009

    2022

  13. [21]

    Y. Cong, S. Khanna, C. Meng, P. Liu, E. Rozi, Y. He, M. Burke, D. Lo- bell, S. Ermon, Satmae: Pre-training transformers for temporal and multi-spectral satellite imagery, Advances in Neural Information Pro- cessing Systems 35 (2022) 197–211

    2022

  14. [22]

    X. Sun, P. Wang, W. Lu, Z. Zhu, X. Lu, Q. He, J. Li, X. Rong, Z. Yang, H. Chang, et al., Ringmo: A remote sensing foundation model with masked image modeling, IEEE Transactions on Geoscience and Remote Sensing 61 (2022) 1–22

    2022

  15. [23]

    J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, L. Fei-Fei, Imagenet: A large-scale hierarchical image database, in: 2009 IEEE conference on computer vision and pattern recognition, Ieee, 2009, pp. 248–255

    2009

  16. [24]

    B. Han, S. Zhang, X. Shi, M. Reichstein, Bridging remote sensors with multisensor geospatial foundation models, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 27852–27862

    2024

  17. [25]

    F. Liu, D. Chen, Z. Guan, X. Zhou, J. Zhu, Q. Ye, L. Fu, J. Zhou, Remoteclip: A vision language foundation model for remote sensing, IEEE Transactions on Geoscience and Remote Sensing (2024)

    2024

  18. [26]

    Z. Wang, R. Prabha, T. Huang, J. Wu, R. Rajagopal, Skyscript: A large and semantically diverse vision-language dataset for remote sens- ing, in: Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, 2024, pp. 5805–5813

    2024

  19. [27]

    U. Mall, C. P. Phoo, M. K. Liu, C. Vondrick, B. Hariharan, K. Bala, Remote sensing vision-language foundation models without annotations via ground remote alignment, arXiv preprint arXiv:2312.06960 (2023)

    2023 arXiv

  20. [28]

    Kuckreja, M

    K. Kuckreja, M. S. Danish, M. Naseer, A. Das, S. Khan, F. S. Khan, Geochat: Grounded large vision-language model for remote sensing, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, 2024, pp. 27831–27840. 34

    2024

  21. [29]

    H. Liu, C. Li, Y. Li, Y. J. Lee, Improved baselines with visual instruction tuning, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 26296–26306

    2024

  22. [30]

    X. Li, C. Li, Z. Tong, A. Lim, J. Yuan, Y. Wu, J. Tang, R. Huang, Campus3d: A photogrammetry point cloud benchmark for hierarchi- cal understanding of outdoor scene, in: Proceedings of the 28th ACM International Conference on Multimedia, 2020, pp. 238–246

    2020

  23. [31]

    Q. Hu, B. Yang, S. Khalid, W. Xiao, N. Trigoni, A. Markham, Sen- saturban: Learning semantics from urban-scale photogrammetric point clouds, International Journal of Computer Vision 130 (2022) 316–343

    2022

  24. [32]

    M. Li, Y. Wu, A. G. Yeh, F. Xue, Hrhd-hk: A benchmark dataset of high-rise and high-density urban scenes for 3d semantic segmentation of photogrammetric point clouds, in: 2023 IEEE international conference on image processing challenges and workshops (ICIPCW), volume 1, 2023, ...

    2023

  25. [33]

    M. Chen, Q. Hu, Z. Yu, H. Thomas, A. Feng, Y. Hou, K. McCul- lough, F. Ren, L. Soibelman, Stpls3d: A large-scale synthetic and real aerial photogrammetry 3d point cloud dataset, arXiv preprint arXiv:2203.09065 (2022)

    2022

  26. [34]

    Hackel, N

    T. Hackel, N. Savinov, L. Ladicky, J. D. Wegner, K. Schindler, M. Polle- feys, Semantic3d. net: A new large-scale point cloud classification benchmark, arXiv preprint arXiv:1704.03847 (2017)

    2017 arXiv

  27. [35]

    Roynard, J.-E

    X. Roynard, J.-E. Deschaud, F. Goulette, Paris-lille-3d: A large and high-quality ground-truth urban point cloud dataset for automatic seg- mentation and classification, The International Journal of Robotics Re- search 37 (2018) 545–557

    2018

  28. [36]

    Behley, M

    J. Behley, M. Garbade, A. Milioto, J. Quenzel, S. Behnke, C. Stachniss, J. Gall, Semantickitti: A dataset for semantic scene understanding of lidar sequences, in: Proceedings of the IEEE/CVF international conference on computer vision, 2019, pp. 9297–9307

    2019

  29. [37]

    W. Tan, N. Qin, L. Ma, Y. Li, J. Du, G. Cai, K. Yang, J. Li, Toronto-3d: A large-scale mobile lidar dataset for semantic segmentation of urban 35 roadways, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops, 2020, pp. 202–203

    2020

  30. [38]

    Rottensteiner, G

    F. Rottensteiner, G. Sohn, M. Gerke, J. D. Wegner, U. Breitkopf, J. Jung, Results of the isprs benchmark on urban object detection and 3d building reconstruction, ISPRS journal of photogrammetry and remote sensing 93 (2014) 256–271

    2014

  31. [39]

    Zolanvari, S

    S. Zolanvari, S. Ruano, A. Rana, A. Cummins, R. E. Da Silva, M. Rah- bar, A. Smolic, Dublincity: Annotated lidar point cloud and its appli- cations, arXiv preprint arXiv:1909.03613 (2019)

    2019 arXiv

  32. [40]

    Z. Ye, Y. Xu, R. Huang, X. Tong, X. Li, X. Liu, K. Luan, L. Hoegner, U. Stilla, Lasdu: A large-scale aerial lidar dataset for semantic labeling in dense urban areas, ISPRS International Journal of Geo-Information 9 (2020) 450

    2020

  33. [41]

    Devlin, Bert: Pre-training of deep bidirectional transformers for lan- guage understanding, arXiv preprint arXiv:1810.04805 (2018)

    J. Devlin, Bert: Pre-training of deep bidirectional transformers for lan- guage understanding, arXiv preprint arXiv:1810.04805 (2018)

    2018 arXiv

  34. [42]

    X. Yu, L. Tang, Y. Rao, T. Huang, J. Zhou, J. Lu, Point-bert: Pre- training 3d point cloud transformers with masked point modeling, in: Proceedings of the IEEE/CVF conference on computer vision and pat- tern recognition, 2022, pp. 19313–19322

    2022

  35. [43]

    K. Fu, P. Gao, S. Liu, L. Qu, L. Gao, M. Wang, Pos-bert: Point cloud one-stage bert pre-training, Expert Systems with Applications 240 (2024) 122563

    2024

  36. [44]

    K. Fu, M. Yuan, S. Liu, M. Wang, Boosting point-bert by multi-choice tokens, IEEE Transactions on Circuits and Systems for Video Technol- ogy 34 (2023) 438–447

    2023

  37. [45]

    J. T. Rolfe, Discrete variational autoencoders, 2017. URL: https:// arxiv.org/abs/1609.02200. arXiv:1609.02200

    2017 arXiv

  38. [46]

    Y. Pang, W. Wang, F. E. Tay, W. Liu, Y. Tian, L. Yuan, Masked autoencoders for point cloud self-supervised learning, in: European conference on computer vision, Springer, 2022, pp. 604–621. 36

    2022

  39. [47]

    H. Liu, M. Cai, Y. J. Lee, Masked discrimination for self-supervised learning on point clouds, in: European Conference on Computer Vision, Springer, 2022, pp. 657–675

    2022

  40. [48]

    Zhang, Z

    R. Zhang, Z. Guo, P. Gao, R. Fang, B. Zhao, D. Wang, Y. Qiao, H. Li, Point-m2ae: multi-scale masked autoencoders for hierarchical point cloud pre-training, Advances in neural information processing systems 35 (2022) 27061–27074

    2022

  41. [49]

    G. Chen, M. Wang, Y. Yang, K. Yu, L. Yuan, Y. Yue, Pointgpt: Auto- regressively generative pre-training from point clouds, Advances in Neu- ral Information Processing Systems 36 (2024)

    2024

  42. [50]

    S. Yan, Y. Yang, Y. Guo, H. Pan, P.-s. Wang, X. Tong, Y. Liu, Q. Huang, 3d feature prediction for masked-autoencoder-based point cloud pre- training, arXiv preprint arXiv:2304.06911 (2023)

    2023 arXiv

  43. [51]

    Zhang, L

    R. Zhang, L. Wang, Y. Qiao, P. Gao, H. Li, Learning 3d representations from 2d pre-trained models via image-to-point masked autoencoders, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 21769–21780

    2023

  44. [52]

    Z. Guo, R. Zhang, L. Qiu, X. Li, P.-A. Heng, Joint-mae: 2d-3d joint masked autoencoders for 3d point cloud pre-training, arXiv preprint arXiv:2302.14007 (2023)

    2023 arXiv

  45. [53]

    Z. Qi, R. Dong, G. Fan, Z. Ge, X. Zhang, K. Ma, L. Yi, Contrast with reconstruct: Contrastive 3d representation learning guided by generative pretraining, in: International Conference on Machine Learning, PMLR, 2023, pp. 28223–28243

    2023

  46. [54]

    G. Hess, J. Jaxing, E. Svensson, D. Hagerman, C. Petersson, L. Svens- son, Masked autoencoder for self-supervised pre-training on lidar point clouds, in: Proceedings of the IEEE/CVF winter conference on appli- cations of computer vision, 2023, pp. 350–359

    2023

  47. [55]

    X. Tian, H. Ran, Y. Wang, H. Zhao, Geomae: Masked geometric target prediction for self-supervised point cloud pre-training, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog- nition, 2023, pp. 13570–13580. 37

    2023

  48. [56]

    H. Yang, T. He, J. Liu, H. Chen, B. Wu, B. Lin, X. He, W. Ouyang, Gd-mae: generative decoder for mae pre-training on lidar point clouds, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 9403–9414

    2023

  49. [57]

    Car´ os, A

    M. Car´ os, A. Just, S. Segu ´ ı, J. Vitri` a, Self-supervised pre-training boosts semantic scene segmentation on lidar data, in: 2023 18th International Conference on Machine Vision and Applications (MV A), IEEE, 2023, pp. 1–6

    2023

  50. [58]

    Zbontar, L

    J. Zbontar, L. Jing, I. Misra, Y. LeCun, S. Deny, Barlow twins: Self- supervised learning via redundancy reduction, in: International confer- ence on machine learning, PMLR, 2021, pp. 12310–12320

    2021

  51. [59]

    de G´ elis, S

    I. de G´ elis, S. Saha, M. Shahzad, T. Corpetti, S. Lef` evre, X. X. Zhu, Deep unsupervised learning for 3d als point clouds change detection, ISPRS Open Journal of Photogrammetry and Remote Sensing 9 (2023) 100044

    2023

  52. [60]

    H. Yang, S. Huang, R. Wang, X. Wang, Self-supervised pre-training for 3d roof reconstruction on lidar data, IEEE Geoscience and Remote Sensing Letters (2024)

    2024

  53. [61]

    Zhang, J

    Y. Zhang, J. Yao, R. Zhang, X. Wang, S. Chen, H. Fu, Havana: Hard negative sample-aware self-supervised contrastive learning for airborne laser scanning point cloud semantic segmentation, Remote Sensing 16 (2024) 485

    2024

  54. [62]

    U. G. Survey, What is 3dep?, 2024. URL: https://www.usgs.gov/ 3d-elevation-program/what-3dep , accessed: 2024-12-02

    2024

  55. [63]

    U. G. Survey, Lidar explorer, 2024. URL: https://apps.nationalmap. gov/lidar-explorer/#/, accessed: 2024-12-02

    2024

  56. [64]

    U. G. Survey, The national map downloader, 2024. URL: https: //apps.nationalmap.gov/downloader/, accessed: 2024-12-02

    2024

  57. [65]

    Cretini, The 3D Elevation Program—Supporting Arkansas’s econ- omy, Technical Report, US Geological Survey, 2023

    C. Cretini, The 3D Elevation Program—Supporting Arkansas’s econ- omy, Technical Report, US Geological Survey, 2023. 38

    2023

  58. [66]

    C. M. Rachol, The 3D Elevation Program—Supporting Michigan’s econ- omy, Technical Report, US Geological Survey, 2024

    2024

  59. [67]

    Fredericks, C

    X. Fredericks, C. Cretini, The 3D Elevation Program—Supporting Florida’s economy, Technical Report, US Geological Survey, 2024

    2024

  60. [68]

    Chirico, J

    P. Chirico, J. DeWitt, S. Bergstresser, Evaluating elevation change thresholds between structure-from-motion dems derived from historical aerial photos and 3dep lidar data, Remote Sensing 12 (2020) 1625

    2020

  61. [69]

    S. Oh, J. Jung, G. Shao, G. Shao, J. Gallion, S. Fei, High-resolution canopy height model generation and validation using usgs 3dep lidar data in indiana, usa, Remote Sensing 14 (2022) 935

    2022

  62. [70]

    C. P. Scott, M. Beckley, M. Phan, E. Zawacki, C. Crosby, V. Nandigam, R. Arrowsmith, Statewide usgs 3dep lidar topographic differencing ap- plied to indiana, usa, Remote Sensing 14 (2022) 847

    2022

  63. [71]

    U. G. Survey, Usgs 3dep lidar point clouds, 2024. URL: https:// registry.opendata.aws/usgs-lidar/, accessed: 2024-12-02

    2024

  64. [72]

    J. R. Anderson, A land use and land cover classification system for use with remote sensor data, volume 964, US Government Printing Office, 1976

    1976

  65. [73]

    M.-R. L. C. C. (MRLC), National land cover database class legend and description, https://www.mrlc.gov/data/legends/ national-land-cover-database-class-legend-and-description ,

  66. [74]

    Pamela, A

    P. Pamela, A. Yukni, S. A. Imam, R. D. Kartiko, The selective causative factors on landslide susceptibility assessment: Case study takengon, aceh, indonesia, in: AIP Conference Proceedings, volume 1987, AIP Publishing, 2018

    1987

  67. [75]

    Chegini, H.-Y

    T. Chegini, H.-Y. Li, L. R. Leung, HyRiver: Hydroclimate Data Re- triever, Journal of Open Source Software 6 (2021) 1–3. doi: 10.21105/ joss.03175

    2021

  68. [76]

    U. G. Survey, Thematic accuracy assessment of nlcd 2019 land cover for the conterminous united states, 39

    2019

  69. [77]

    Melekhov, A

    I. Melekhov, A. Umashankar, H.-J. Kim, V. Serkov, D. Argyle, Eclair: A high-fidelity aerial lidar dataset for semantic segmentation, in: Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 7627–7637

    2024

  70. [78]

    URL: https://www.usgs.gov/publications/ thematic-accuracy-assessment-nlcd-2019-land-cover-conterminous-united-states , accessed: 2024-12-02

    2019

  71. [79]

    Sithole, G

    G. Sithole, G. Vosselman, Experimental comparison of filter algorithms for bare-earth extraction from airborne laser scanning point clouds, IS- PRS journal of photogrammetry and remote sensing 59 (2004) 85–101

    2004

  72. [80]

    Graves, S

    S. Graves, S. Marconi, Idtrees 2020 competition data (2020)

    2020

  73. [81]

    C. Choy, J. Gwak, S. Savarese, 4d spatio-temporal convnets: Minkowski convolutional neural networks, in: Proceedings of the IEEE/CVF con- ference on computer vision and pattern recognition, 2019, pp. 3075– 3084

    2019

  74. [82]

    Graham, M

    B. Graham, M. Engelcke, L. Van Der Maaten, 3d semantic segmentation with submanifold sparse convolutional networks, in: Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 9224–9232

    2018

  75. [83]

    Ronneberger, P

    O. Ronneberger, P. Fischer, T. Brox, U-net: Convolutional networks for biomedical image segmentation, in: Medical image computing and computer-assisted intervention–MICCAI 2015: 18th international con- ference, Munich, Germany, October 5-9, 2015, proceedings, part III 18, Spri...

    2015

  76. [84]

    N. Qin, X. Hu, H. Dai, Deep fusion of multi-view and multimodal representation of als point cloud for 3d terrain scene recognition, ISPRS journal of photogrammetry and remote sensing 143 (2018) 205–212

    2018

  77. [85]

    Thomas, C

    H. Thomas, C. R. Qi, J.-E. Deschaud, B. Marcotegui, F. Goulette, L. J. Guibas, Kpconv: Flexible and deformable convolution for point clouds, 40 in: Proceedings of the IEEE/CVF international conference on computer vision, 2019, pp. 6411–6420

    2019

  78. [86]

    C. R. Qi, L. Yi, H. Su, L. J. Guibas, Pointnet++: Deep hierarchical feature learning on point sets in a metric space, Advances in neural information processing systems 30 (2017)

    2017

  79. [87]

    S. Yoo, Y. Jeong, M. Jameela, G. Sohn, Human vision based 3d point cloud semantic segmentation of large-scale outdoor scenes, in: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 6577–6586

    2023

  80. [88]

    Q. Hu, B. Yang, L. Xie, S. Rosa, Y. Guo, Z. Wang, N. Trigoni, A. Markham, Randla-net: Efficient semantic segmentation of large- scale point clouds, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 11108–11117

    2020

  81. [89]

    C. J. Reed, R. Gupta, S. Li, S. Brockman, C. Funk, B. Clipp, K. Keutzer, S. Candido, M. Uyttendaele, T. Darrell, Scale-mae: A scale-aware masked autoencoder for multiscale geospatial representation learning, in: Proceedings of the IEEE/CVF International Conference on Com- pute...

    2023

  82. [90]

    Tukra, F

    S. Tukra, F. Hoffman, K. Chatfield, Improving visual representation learning through perceptual understanding, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 14486–14495

    2023

  83. [92]

    Hinton, Distilling the knowledge in a neural network, arXiv preprint arXiv:1503.02531 (2015)

    G. Hinton, Distilling the knowledge in a neural network, arXiv preprint arXiv:1503.02531 (2015). 41

    2015 arXiv

  84. [2024]

    Accessed: 2024-12-03

    2024

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.