Pith. sign in

REVIEW 5 major objections 6 minor 1 cited by

Online Adaptive Traversability Estimation through Interaction for Unstructured, Densely Vegetated Environments

T0 review · 5 major / 6 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read A robot can train its own terrain model in 8 minutes, lidar-only, while pushing through dense bush, and then navigate that same forest on its own.

desk verdict The paper delivers a genuine first — online lidar-only traversability training on a robot in under eight minutes — and the main caveats are the self-admitted steady-state fusion assumption and the single-run headline. read the letter →

arxiv 2502.01987 v1 pith:5GWKOZF4 submitted 2025-02-04 cs.RO

Fabio A. Ruetz , Nicholas Lawrance , Emili Hernández , Paulo V. K. Borges , Thierry Peynot This is my paper · ORCID
classification cs.RO
keywords traversabilityestimationonlineadaptationlidar-onlyself-supervisedlearningvoxelmapvegetatedenvironmentssparsegraphfieldrobotics
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 claims that a lidar-only traversability estimator can be trained from scratch on a robot in the field, using only self-supervised labels from the robot physically interacting with vegetation, and that the resulting model enables closed-loop point-to-point navigation in a dense forest. The core idea is that a 3D probabilistic voxel map, fused with collision states, can be maintained as a sparse graph so that fresh training data accumulates without storing a global map. In a demonstration, less than eight minutes of operator-guided interaction produced a model scoring 0.63 Matthews Correlation Coefficient on a held-out forest scene, and that model safely drove the robot along a waypoint trajectory. The paper also compares four online training strategies and recommends continuous adaptation from a pre-trained base model as the most consistent.

What carries the argument

The central object is the Online Data Graph (OGraph), a sparse graph whose nodes hold small, local 3D probabilistic voxel maps built from lidar measurements and robot collision experience. Each voxel stores distributions for NDT-OM occupancy (a 3D Gaussian of endpoint geometry), NDT-TM permeability (hit/miss counts), laser intensity statistics, and multi-return counts; these distributions directly feed a sparse-convolutional UNet. The graph collapses many robot poses into a sparse representation, and the key update rule is the steady-state assumption: with sufficient measurements, the probabilistic voxel and collision maps converge, so newer measurements overwrite older ones. That rule allows the system to keep training data current without maintaining a dense global map, which is what makes real-time training on a 25W GPU possible.

What would settle it

Replay a single environment twice from different viewpoints or at different times, and compare the voxel distributions produced by the OGraph update; if a substantial fraction of voxels change their traversability belief with each new pass rather than converging to a stable value, the steady-state assumption underlying the online data generation is violated.

Watch

Extended reading notes

Core claim

The central claim is that a randomly initialised sparse-convolutional network for voxel-wise traversability estimation can be trained online, on the robot, entirely from experience labels gathered in situ, and reach accuracy comparable to offline-trained models. Concretely, the online-trained model achieved an MCC of 0.63 on test scene #9 and enabled the robot to navigate point-to-point through underbrush, grass, and brambles. The authors argue that this shows online adaptation with probabilistic 3D voxel representations is feasible, that no hand-labelling is needed for deployment in a novel dense environment, and that a purely geometric lidar-based method can compete with image-based self-supervised approaches that have so far dominated this problem.

Load-bearing premise

The method assumes that the probabilistic voxel and collision maps converge to a steady state once enough measurements have been seen, so newer data can safely overwrite older data; if a voxel's distribution keeps changing with viewpoint or time instead of settling, the online training labels become inconsistent and the learned model degrades.

Editorial extensions

If this is right

  • If the central claim holds, a robot can be adapted to a novel vegetated environment in minutes during deployment, removing the need for offline hand-labelled data collection before every new site.
  • The comparison of training strategies suggests that a pre-trained base model combined with continuous adaptation (BM 1, CA 1) gives the most stable and best-performing online traversal estimates, while retraining from scratch at each cycle oscillates in accuracy.
  • The finding that a model trained on dense forest data generalises well to a structurally different industrial environment implies that training on complex, varied vegetation may transfer to simpler settings without retraining.
  • The 0.63 MCC achieved by the online-trained model matches the offline LfE-only baseline, indicating that self-supervised experience alone, gathered in less than eight minutes, can reach the same accuracy as offline training on post-processed data.
  • Real-time training on a 25W GPU makes the approach practical for field robots with limited onboard compute, not just research platforms with desktop GPUs.

Reading between the lines

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

  • A natural extension the authors do not pursue is fully autonomous continual learning: if the OGraph keeps updating during nominal operation, the model could keep adapting without an operator, provided collisions are detected reliably enough to label new data.
  • The steady-state assumption could be tested and possibly replaced by an explicit uncertainty estimate per voxel distribution, which would let the system weight newer versus older measurements based on observed convergence rather than fixed overwriting.
  • The demonstrated generalisation from dense forest to industrial scenes suggests that a shared geometry-based feature space may underpin traversability across very different environments; if true, carefully chosen dense-forest data could serve as a broader pre-training set for off-road navigation.
  • The method is currently demonstrated on a tracked vehicle; adapting it to wheeled or legged platforms would require recalibrating the collision bounding box, but the voxel-distribution machinery itself is platform-agnostic.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 6 minor

Summary. This paper presents an online adaptive traversability estimation method for lidar-only ground robots in vegetated terrain. The method extends the authors' prior ForestTrav voxel representation with (i) a Bayesian collision-state mapping that self-labels voxels through robot-environment interaction, (ii) a sparse Online Data Graph (OGraph) that fuses temporally evolving probabilistic voxel maps with collision labels, and (iii) an online training module that retrains or fine-tunes a sparse-convolutional U-Net on the robot. The authors report that a randomly initialised model trained in situ in less than 8 minutes achieves MCC 0.63 on a held-out dense-forest test scene, compare four training strategies in a replay experiment, and demonstrate point-to-point navigation against several baselines in two forest locations.

Significance. If the reported results hold, the work is a useful step for field robotics: it shows that self-supervised learning-from-experience data can be generated and used to train a 3D voxel traversability model entirely onboard a resource-constrained platform, and it provides a practical comparison of training strategies. The controlled replay experiment, the navigation comparison against NavStack and ForestTrav variants, and the open-source code release are valuable assets. The central quantitative claim, however, is currently supported by a single real-time run and depends on an unvalidated steady-state assumption in the OGraph update, so the contribution in its present form is mainly a demonstration rather than a fully supported performance claim.

major comments (5)
  1. [Section IV-D] The headline result (MCC 0.63 after less than 8 minutes) is a single real-time run, with no repeated runs, no seed variation, and no confidence interval. This is particularly important because Table 2 reports 0.69 ± 0.022 for the offline pure-LfE model, so the online value is more than two standard deviations below that baseline; the claim that the online model is 'comparable' therefore rests on a single point estimate. Please either report multiple online runs or explicitly reframe the result as a single-case demonstration and adjust the abstract and conclusion wording accordingly.
  2. [Section III-D and Section V] The OGraph overwrite rule assumes that the probabilistic voxel and collision maps converge to a steady state, and the Discussion admits that convergence introspection is a current limitation. Because the online training pairs Bayesian-filtered collision labels with feature vectors that may not be converged after the 40 s training cycles, the quality of the self-supervised data, and hence the reported MCC, depends on this assumption. Please provide evidence on distribution convergence (e.g., voxel revisit counts and distribution drift over the 8-minute run) and an ablation on the update rule (newest versus averaged versus first measurements) using the replayed online data set.
  3. [Section IV-E and Figure 8] The replay experiment does not state the number of independent runs used to produce the means and the shaded one-standard-deviation bands, nor does it define how seeds or random initialisations are handled. The 'pre-determined scaling values' used for randomly initialised models are also not specified. Without these details, the claimed ranking of the four training strategies (e.g., that BM 1/CA 1 has the lowest variance) is not reproducible or statistically assessable.
  4. [Section IV-F and Table 5] The navigation comparison reports one trial per method at each location. Because the learned model and the hybrid-A* planner can be stochastic, and because the success criterion includes operator interventions, a single trial per method cannot support robust claims about 'safe navigation' or relative method robustness. The authors should either add repeated trials with a clear protocol for when an intervention counts as a failure or restrict the claims to qualitative demonstration.
  5. [Section III-B2] The collision bounding-box extension (0.1 m behind the front of the chassis and 0.2 m beyond it at 0.1 m voxel resolution) is stated to be 'found heuristically,' and the collision update uses 'a fixed probability' whose value is not reported. These parameters directly control which voxels are labelled non-traversable and therefore what the self-supervised model learns; a small change could shift the resulting MCC substantially. A sensitivity analysis over the bounding-box extension and the collision update probability should be added, and the missing probability value should be reported.
minor comments (6)
  1. [Section IV-E] The text refers to '(BM 0, FT 1)', but the flags defined in that section are BM and CA; this should be '(BM 0, CA 1)'.
  2. [Table 1] The Scene #12 row reports '42' in the NTR HL column; this should be '0.42' to be consistent with the other percentage entries.
  3. [Table 4] The BM:SPARSE, DENSE row contains '0.80 0.79)' with an unmatched parenthesis, and the BM:DENSE, SPARSE row has '0.67 (0. 4)', which appears to be a typo for '0.84'.
  4. [Section III-D] The phrase 'described in Section 2' should reference the actual subsection, 'Section III-B2', for the collision-map generation.
  5. [Section III-G2 and Section IV-C4] The text says 40 epochs for incrementally-trained models but later states that post-processed fine-tuning was limited to 150 epochs 'to make sure the results were comparable to the online case'; please clarify which epoch count applies to the online cycles in Figure 8.
  6. [Figure 8] The x-axis is labelled 'time' without units; the caption should specify minutes or seconds.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the online-trained model is evaluated on a held-out test scene, and the self-citations to prior work are not load-bearing.

full rationale

The paper's central claim is that a randomly initialised model trained online with self-supervised LfE data reaches an MCC of 0.63 on held-out forest test scene #9. The online training data was collected 'within a new area' distinct from scene #9, and the paper states that 'The test data set remains scene #9 from the previously established data set, allowing for a comparison to all other experiments.' The evaluation is therefore not performed on data used to train the online model. No parameter is fitted to the test set; the LfE labels are generated by robot interaction through the collision map, and the model is trained from scratch (BM 0) or fine-tuned from a base model, with the held-out test scene providing independent labels. The OGraph steady-state overwrite rule is an unvalidated modelling assumption, but it is not circular: it does not define the reported MCC or the navigation outcome, and it is explicitly acknowledged as a limitation. The architecture and voxel features are adopted from the authors' prior ForestTrav work [19], but that work is external and supplies the representation; the online adaptation contribution is validated against a held-out test set and against baseline methods, so the self-citations do not carry the derivation. Heuristic costmap parameters are stated as expert heuristics rather than fitted predictions, and the navigation experiments compare multiple methods without using the online test scene to tune the method. Overall, no prediction in the paper reduces by construction to an input, and no load-bearing self-citation chain is present. The score of 2 reflects only the presence of minor, non-load-bearing self-citations to prior work by the same authors.

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

The online learning pipeline relies on several stated domain assumptions and hand-chosen parameters. The most important are the steady-state convergence of voxel distributions, the reliability of collision-based self-labelling, and the sufficiency of the lidar feature set inherited from ForestTrav. The heuristic collision bounding box and costmap constants are free parameters that affect reported performance.

free parameters (4)
  • Collision bounding box extension = 0.1 m behind front plate, 1-2 voxels (0.2 m) beyond chassis
    Section III-B2: 'The distance threshold values were found heuristically based on the environment and the voxel resolution.'
  • Costmap heuristic parameters = N=10, zmax=0.5, NAdj=5, 5x5 kernel, lambda_tau=0.3
    Section III-F3: 'These values were determined heuristically by an expert with knowledge of the environment, the robot dimensions and observation in the field.'
  • OGraph node parameters = dnode=0.5 m, rmax=2 m, zmin=-0.5 m, zmax=0.8 m
    Section III-D and III-G3: values guided by robot dimensions and prior work; they affect the training data distribution.
  • Training cycle interval = delta_t = 40 s
    Section III-G2: 'The empirically found duration of training cycles is a balance between generating sufficient new updates, allowing for the processing of the updates and accommodating overheads for the continuation of the training.'
assumptions (4)
  • domain assumption Voxel distributions are independent of adjacent voxels
    Section III-B: 'Voxel distributions are assumed independent of the adjacent voxels.' This underpins the per-voxel feature representation.
  • domain assumption The lidar feature set (occupancy, NDT Gaussian, permeability, intensity, multi-return) is sufficient to discriminate traversability in vegetation
    Carried over from prior ForestTrav work [19]; the paper relies on this feature set without re-deriving it.
  • ad hoc to paper Probabilistic voxel and collision maps converge to a steady state with sufficient measurements, allowing newest measurements to overwrite older ones
    Section III-D: load-bearing for OGraph data generation, but not validated empirically.
  • domain assumption The robot collision state recorded by the operator is a reliable indicator of voxel-level traversability labels derived from a bounding box
    Section III-B2: the self-labelling mechanism assumes collisions correspond to non-traversable voxels in the bounding box, which may be noisy.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Online Adaptive Traversability Estimation through Interaction for Unstructured, Densely Vegetated Environments." pith.science (2026). https://pith.science/paper/5GWKOZF4

@misc{pith2026250201987,
  author       = {Pith},
  title        = {Pith review of: Online Adaptive Traversability Estimation through Interaction for Unstructured, Densely Vegetated Environments},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5GWKOZF4}},
  note         = {Machine review of arXiv:2502.01987}
}
read the original abstract

Navigating densely vegetated environments poses significant challenges for autonomous ground vehicles. Learning-based systems typically use prior and in-situ data to predict terrain traversability but often degrade in performance when encountering out-of-distribution elements caused by rapid environmental changes or novel conditions. This paper presents a novel, lidar-only, online adaptive traversability estimation (TE) method that trains a model directly on the robot using self-supervised data collected through robot-environment interaction. The proposed approach utilises a probabilistic 3D voxel representation to integrate lidar measurements and robot experience, creating a salient environmental model. To ensure computational efficiency, a sparse graph-based representation is employed to update temporarily evolving voxel distributions. Extensive experiments with an unmanned ground vehicle in natural terrain demonstrate that the system adapts to complex environments with as little as 8 minutes of operational data, achieving a Matthews Correlation Coefficient (MCC) score of 0.63 and enabling safe navigation in densely vegetated environments. This work examines different training strategies for voxel-based TE methods and offers recommendations for training strategies to improve adaptability. The proposed method is validated on a robotic platform with limited computational resources (25W GPU), achieving accuracy comparable to offline-trained models while maintaining reliable performance across varied environments.

Figures

Figures reproduced from arXiv: 2502.01987 by the authors.

Figure 1
Figure 1. FIGURE 1: The overview of the online TE adaptation method. The [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. FIGURE 2: Visualisation of the Learning from Experience [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. FIGURE 3: The Left image shows a 3D traversability [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: FIGURE 4: Top left: Shows the overview of the environment [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]
Figure 5
Figure 5. Figure 5: FIGURE 5: Comparison of two ensembles of models trained on either the dense data set in the top row (A, B, C) or the [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: FIGURE 6: Overview of the navigation in a forest environ [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]
Figure 7
Figure 7. Figure 7: FIGURE 7: Visualisation scenes from a successful online point-to-point navigation using the online learnt TE model in [PITH_FULL_IMAGE:figures/full_fig_p015_7.png]
Figure 8
Figure 8. Figure 8: FIGURE 8: Comparison of the performance of different models over the incremental online adaptation for the four different [PITH_FULL_IMAGE:figures/full_fig_p016_8.png]
Figure 9
Figure 9. Figure 9: FIGURE 9: Qualitative examples of the methods in the target environment. Blue blocks are for all methods at location 1 [PITH_FULL_IMAGE:figures/full_fig_p017_9.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Adversarial Stress Testing of Role-Playing Language Agents using Multi-Agent Evaluation

    cs.AI 2026-08 conditional novelty 5.0 of 10

    A multi-agent adversarial evaluation platform with six progressive attack strategies shows that role-playing LLMs degrade under sustained pressure, with automated judging correlating with human ratings.

Reference graph

Works this paper leans on

29 extracted references · 23 canonical work pages · cited by 1 Pith paper

  1. [1]

    Fast traversability estimation for wild visual navigation,

    J. Frey, M. Mattamala, N. Chebrolu, C. Cadena, M. Fallon, and M. Hutter, “Fast traversability estimation for wild visual navigation,” arXiv preprint arXiv:2305.08510 , 2023

    arXiv 2023

  2. [2]

    Learning-based methods of percep- tion and navigation for ground vehicles in unstructured environments: A review,

    D. C. Guastella and G. Muscato, “Learning-based methods of percep- tion and navigation for ground vehicles in unstructured environments: A review,” Sensors, vol. 21, no. 1, p. 73, 2021

    work page 2021

  3. [3]

    Scene understanding for a high-mobility walking robot,

    D. M. Bradley, J. K. Chang, D. Silver et al., “Scene understanding for a high-mobility walking robot,” in IROS, 2015

    work page 2015

  4. [4]

    Self- supervised 3d traversability estimation with proxy bank guidance,

    J. Bae, J. Seo, T. Kim, H.-G. Jeon, K. Kwak, and I. Shim, “Self- supervised 3d traversability estimation with proxy bank guidance,” IEEE Access, vol. 11, pp. 51 490–51 501, 2023. 18 VOLUME , <Society logo(s) and publication title will appear here. >

    work page 2023

  5. [5]

    Where should I walk? predicting terrain properties from images via Self-Supervised learning,

    L. Wellhausen, A. Dosovitskiy, R. Ranftl et al. , “Where should I walk? predicting terrain properties from images via Self-Supervised learning,” IEEE RAL, vol. 4, no. 2, 2019

    work page 2019

  6. [6]

    BADGR: An autonomous self- supervised learning-based navigation system,

    G. Kahn, P. Abbeel, and S. Levine, “BADGR: An autonomous self- supervised learning-based navigation system,” IEEE RAL, vol. 6, no. 2, 2021

    work page 2021

  7. [7]

    Seeing Through the Grass: Semantic Pointcloud Filter for Support Surface Learning

    A. Li, C. Yang, J. Frey, J. Lee, C. Cadena, and M. Hutter, “Seeing through the grass: Semantic pointcloud filter for support surface learning,” arXiv preprint arXiv:2305.07995 , 2023

    work page Pith review arXiv 2023

  8. [8]

    Roadrunner–learning traversability estimation for autonomous off-road driving,

    J. Frey, S. Khattak, M. Patel, D. Atha, J. Nubert, C. Padgett, M. Hutter, and P. Spieler, “Roadrunner–learning traversability estimation for autonomous off-road driving,” arXiv preprint arXiv:2402.19341, 2024

    arXiv 2024

Show all 29 references
  1. [9]

    Adaptive robot traversability estimation based on self- supervised online continual learning in unstructured environments,

    H.-S. Yoon, J.-H. Hwang, C. Kim, E. I. Son, S.-W. Yoo, and S.- W. Seo, “Adaptive robot traversability estimation based on self- supervised online continual learning in unstructured environments,” IEEE Robotics and Automation Letters , 2024

    2024

  2. [10]

    Terrain traversability analysis methods for unmanned ground vehicles: A survey,

    P. Papadakis, “Terrain traversability analysis methods for unmanned ground vehicles: A survey,” Engineering Applications of Artificial Intelligence, vol. 26, no. 4, 2013

    2013

  3. [11]

    Heterogeneous ground and air platforms, homogeneous sensing: Team CSIRO Data61’s approach to the DARPA Subterranean Challenge,

    N. Hudson, F. Talbot, M. Cox, and other, “Heterogeneous ground and air platforms, homogeneous sensing: Team CSIRO Data61’s approach to the DARPA Subterranean Challenge,” Field Robotics J. , 2022

    2022

  4. [12]

    Cerberus: Autonomous legged and aerial robotic exploration in the tunnel and urban circuits of the darpa subterranean challenge,

    M. Tranzatto, F. Mascarich, L. Bernreiter, C. Godinho, M. Camurri, S. Khattak, T. Dang, V . Reijgwart, J. Loeje, D. Wisthet al., “Cerberus: Autonomous legged and aerial robotic exploration in the tunnel and urban circuits of the darpa subterranean challenge,” arXiv preprint ar...

    2022 arXiv

  5. [13]

    Online adaptive rough-terrain naviga- tion vegetation,

    C. Wellington and A. Stentz, “Online adaptive rough-terrain naviga- tion vegetation,” in IEEE International Conference on Robotics and Automation, 2004. Proceedings. ICRA ’04. 2004 , vol. 1, Apr. 2004, pp. 96–101

    2004

  6. [14]

    Rspmp: Real-time semantic perception and motion planning for autonomous navigation of unmanned ground vehicle in off-road environments,

    D. Chen, M. Zhuang, X. Zhong, W. Wu, and Q. Liu, “Rspmp: Real-time semantic perception and motion planning for autonomous navigation of unmanned ground vehicle in off-road environments,” Applied Intelligence, vol. 53, no. 5, pp. 4979–4995, 2023

    2023

  7. [15]

    Ros: an open-source robot operating system,

    M. Quigley, K. Conley, B. Gerkey, J. Faust, T. Foote, J. Leibs, R. Wheeler, A. Y . Ng et al. , “Ros: an open-source robot operating system,” in ICRA workshop on open source software , vol. 3, no. 3.2. Kobe, Japan, 2009, p. 5

    2009

  8. [16]

    STEP: Stochastic traversability evaluation and planning for risk-aware off-road navigation,

    D. D. Fan, K. Otsu, Y . Kubo et al. , “STEP: Stochastic traversability evaluation and planning for risk-aware off-road navigation,” arXiv preprint arXiv:2103.02828, 2021

    2021 arXiv

  9. [17]

    Normal distributions transform traversability maps: LIDAR-only approach for traversability mapping in outdoor environments,

    J. Ahtiainen, T. Stoyanov, and J. Saarinen, “Normal distributions transform traversability maps: LIDAR-only approach for traversability mapping in outdoor environments,” Journal of Field Robotics, vol. 34, no. 3, 2017

    2017

  10. [18]

    Forest traversability mapping (FTM): Traversability estimation using 3D voxel-based normal distributed transform to enable forest naviga- tion,

    F. Ruetz, P. Borges, N. Suenderhauf, E. Hern ´andez, and T. Peynot, “Forest traversability mapping (FTM): Traversability estimation using 3D voxel-based normal distributed transform to enable forest naviga- tion,” in IROS, 2022

    2022

  11. [19]

    ForestTrav: 3D lidar-only forest traversability estimation for au- tonomous ground vehicles,

    F. Ruetz, N. Lawrance, E. Hern ´andez, P. Borges, and T. Peynot, “ForestTrav: 3D lidar-only forest traversability estimation for au- tonomous ground vehicles,” IEEE Access, 2024

    2024

  12. [20]

    Traversabil- ity classification using unsupervised on-line visual learning for outdoor robot navigation,

    D. Kim, J. Sun, S. M. Oh, J. M. Rehg, and A. F. Bobick, “Traversabil- ity classification using unsupervised on-line visual learning for outdoor robot navigation,” in Proceedings 2006 IEEE International Conference on Robotics and Automation, 2006. ICRA 2006. IEEE, 2006, pp. 518–525

    2006

  13. [21]

    Learning long-range vision for autonomous off-road driving,

    R. Hadsell, P. Sermanet, J. Ben, A. Erkan, M. Scoffier, K. Kavukcuoglu, U. Muller, and Y . LeCun, “Learning long-range vision for autonomous off-road driving,” Journal of Field Robotics , vol. 26, no. 2, pp. 120–144, 2009

    2009

  14. [22]

    OctoMap: An efficient probabilistic 3d mapping framework based on octrees,

    A. Hornung, K. M. Wurm, M. Bennewitz, C. Stachniss, and W. Bur- gard, “OctoMap: An efficient probabilistic 3d mapping framework based on octrees,” Autonomous robots, vol. 34, pp. 189–206, 2013

    2013

  15. [23]

    Normal distributions transform occupancy maps: Appli- cation to large-scale online 3d mapping,

    J. Saarinen, H. Andreasson, T. Stoyanov, J. Ala-Luhtala, and A. J. Lilienthal, “Normal distributions transform occupancy maps: Appli- cation to large-scale online 3d mapping,” in 2013 IEEE international conference on robotics and automation. IEEE, 2013, pp. 2233–2238

    2013

  16. [24]

    Wildcat: On- line continuous-time 3D lidar-inertial SLAM,

    M. Ramezani, K. Khosoussi, G. Catt et al. , “Wildcat: On- line continuous-time 3D lidar-inertial SLAM,” arXiv preprint arXiv:2205.12595, 2022

    2022 arXiv

  17. [25]

    End-to-end path estimation and automatic dataset generation for robot navigation in plant-rich environments,

    Y . Uzawa, S. Matsuzaki, H. Masuzawa, and J. Miura, “End-to-end path estimation and automatic dataset generation for robot navigation in plant-rich environments,” in International Conference on Intelligent Autonomous Systems. Springer, 2022, pp. 272–284

    2022

  18. [26]

    TorchSparse: Efficient Point Cloud Inference Engine,

    H. Tang, Z. Liu, X. Li, Y . Lin, and S. Han, “TorchSparse: Efficient Point Cloud Inference Engine,” in Conference on Machine Learning and Systems (MLSys) , 2022

    2022

  19. [27]

    Rational decisions,

    I. J. Good, “Rational decisions,” Journal of the Royal Statistical Society: Series B (Methodological), vol. 14, no. 1, pp. 107–114, 1952

    1952

  20. [28]

    Practical search techniques in path planning for autonomous driving,

    D. Dolgov, S. Thrun, M. Montemerlo, and J. Diebel, “Practical search techniques in path planning for autonomous driving,” in Proceedings of the First International Symposium on Search Techniques in Artificial Intelligence and Robotics (STAIR-08) . Chicago, USA: AAAI, June 2008

    2008

  21. [29]

    Automatic differen- tiation in pytorch,

    A. Paszke, S. Gross, S. Chintala, G. Chanan, E. Yang, Z. DeVito, Z. Lin, A. Desmaison, L. Antiga, and A. Lerer, “Automatic differen- tiation in pytorch,” in Autodiff Workshop, NIPS 2017 , 2017. Fabio A. Ruetz received his B.S and M.S de- grees in Mechanical and Process Enginee...

    2007

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.