Pith. sign in

REVIEW 3 major objections 5 minor 30 references

Bio-Inspired Hybrid Map: Spatial Implicit Local Frames and Topological Map for Mobile Cobot Navigation

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A hybrid map that pairs learned 3D local frames with a topological factor graph lets a mobile robot navigate with local accuracy near dense neural SLAM while planning routes five times faster.

desk verdict A plausible hybrid mapping/navigation integration whose central precision claim is undermined by comparing per-local-frame RMSE against global ATE baselines. read the letter →

arxiv 2507.04649 v1 pith:MCBL5FT6 submitted 2025-07-07 cs.RO

Tuan Dang , Manfred Huber This is my paper · ORCID
classification cs.RO
keywords mobilerobotnavigationspatial-implicitlocalframestopologicalmapfactorgraphsigneddistancefunctionself-supervisedlearningRRT*continual
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 argues that a mobile robot can navigate unfamiliar indoor spaces without building one globally consistent metric map. The robot instead learns short-term 'spatial-implicit local frames'—sparse 3D points that carry learned neural features encoding signed distance to surfaces—and connects those frames in a topological factor-graph map that stores only relative poses between frames. On the three TUM RGB-D sequences the authors report ATE RMSE of 1.61 cm, 1.54 cm, and 1.99 cm, below the iMAP and NICE-SLAM numbers on all three and below ESLAM on two of three, while the RRT*-based planner produces routes 5x faster (22.1 ms vs 110.6 ms) and 50% shorter (4.5 m vs 6.8 m) than the baseline. The paper's thesis is that local precision and global efficiency can come from two different map representations, with a self-supervised neural surface model doubling as localizer, obstacle map, and source of learned features.

What carries the argument

The load-bearing object is the spatial-implicit local frame: a sparse, semantically labeled point cloud in which each point carries a 3D coordinate, a learned feature vector from the neural network $F_\theta$, and a semantic label. The network is trained as a signed distance function regressor with binary cross-entropy and Eikonal regularization losses plus an elastic weight consolidation term to prevent catastrophic forgetting during continual learning. Pose recovery inside a frame is the critical mechanism: the motion model is set to 1, and the camera pose is found by Levenberg-Marquardt optimization of the SDF residual $r = F_\theta(p, c, f)$ over the Lie-algebra parameter $\xi$ (Eqs. 13–18). The topological factor graph then records arcs $g(LF_i, LF_j, o_k)$ carrying relative poses between frames, and the navigation algorithm is an RRT* variant that samples along a goal vector plus alternative rays separated by a small angle $\alpha$, using local frames for obstacle awareness and the topology for global guidance.

What would settle it

Compute the standard full-trajectory ATE on the TUM fr2/xyz sequence by aligning the complete estimated trajectory to ground truth instead of averaging per-local-frame RMSE; if the drift between local frames that the paper concedes accumulates, the global error will exceed the reported 1.54 cm, showing that the per-frame metric hides the accumulation.

Watch

Extended reading notes

Core claim

The central claim is that decomposing the world into spatial-implicit local frames, each a sparse set of semantically labeled points with learned features, and joining them through a topological factor graph gives a robot both the local accuracy of dense neural SLAM and the global efficiency of topological navigation. Each local frame is built by training a neural signed-distance-field model $F_\theta$ on sampled points along rays from the camera, extracting per-point latent features by backpropagation, and registering subsequent observations against the frame with Levenberg-Marquardt minimization of SDF residuals, with the motion model set to 1. The authors report approximate per-frame ATE RMSE of 1.61 cm, 1.54 cm, and 1.99 cm on TUM fr1/desk, fr2/xyz, and fr3/office, and a navigation planner that samples along a goal vector and its neighboring rays, guided by the topological map, which runs in 22.1 ms and yields a 4.5 m route versus 110.6 ms and 6.8 m for the baseline RRT*. They also note some drift between local frames, which they argue is mitigated when the robot relocalizes within a frame using the spatial-implicit key points.

Load-bearing premise

The system assumes that a robot can recover its pose purely by registering each new RGB-D view against the learned signed-distance surface of the current local frame, with the motion model set to 1 and no odometry, inertial, or loop-closure corrections, so any registration error flows into the local frame, the topological arc, and the navigation plan.

Editorial extensions

If this is right

  • A mobile robot can maintain a usable map at sensor rate (about 25 Hz) without a globally consistent metric reconstruction, as long as each local frame is individually accurate.
  • Planning cost scales with the topological graph rather than the full metric map, which is what makes the reported 5x faster planning (22.1 ms) possible.
  • Continual self-supervised training on every fifth observation keeps the surface model current and lets the robot adapt to new environments without labeled data.
  • The same neural surface model serves as map builder, localizer, and obstacle map, so no separate localization front-end is needed.

Reading between the lines

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

  • Reintroducing an odometry or inertial prior into the pose-update step would probably reduce the inter-frame drift the paper concedes, and full-trajectory ATE would then better match the reported per-frame numbers.
  • The factor-graph topology is a natural fit for multi-session or lifelong mapping: new local frames could be added as nodes, and arcs relinked, without retraining the surface model.
  • The 50% shorter route result depends on environment geometry; a systematic comparison in a shared maze-like arena with forced detours would tell whether the goal-vector sampling or the topological guidance is the main source of the saving.
  • The inverse-distance feature interpolation in Eq. 6 could be replaced with a density-aware weighting; a testable prediction is that local-frame accuracy improves in sparse regions where the current interpolation has few neighbors.
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

3 major / 5 minor

Summary. The paper proposes a bio-inspired hybrid mapping and navigation system for mobile cobots, combining spatial-implicit local frames (3D points with learned neural features and semantic labels) with a topological factor-graph map. A neural network is trained self-supervisely to represent surfaces as signed distance functions (SDFs), camera pose is obtained by Levenberg-Marquardt registration of observed point clouds against the learned SDF with the motion model set to 1, and navigation is performed by an RRT*-based planner guided by the topological map. The authors evaluate on TUM RGB-D sequences and in-lab experiments, reporting local-frame ATE RMSE values of 1.61 cm, 1.54 cm, and 1.99 cm on fr1/desk, fr2/xyz, and fr3/office, and a navigation comparison where their planner is 5x faster and 50% shorter than baseline RRT*. The paper also provides runtime analysis and an open-source code repository.

Significance. If the central claims hold, the hybrid representation is a meaningful step toward scalable neural mapping for navigation, avoiding the global consistency cost of dense metric SLAM while retaining local accuracy through learned implicit surfaces. The paper's strengths include a clearly described self-supervised continual-learning pipeline, open-source code, a real-robot demonstration, and a concrete runtime breakdown on GPU/CPU. However, the evaluation currently does not establish the headline precision claim because the reported metric is not commensurable with the baselines, and the navigation comparison rests on a single anecdotal run. The core idea is defensible and worth pursuing, but the evidence needs substantial strengthening.

major comments (3)
  1. [Section V-C, Table II] The precision comparison is not on a common basis. The paper computes an 'approximate RMSE' over local frames, while iMAP, NICE-SLAM, and ESLAM report standard ATE RMSE over full trajectories. Because per-local-frame errors exclude inter-frame drift, and the text explicitly concedes 'some drift between the local frames,' the values 1.61 cm, 1.54 cm, and 1.99 cm are not directly comparable to the baseline numbers. The authors should report full-trajectory ATE RMSE, or at minimum a per-frame ATE that includes the relative-pose errors between consecutive local frames, to support the claim of 'improved or competitive precision.'
  2. [Section IV-A, Eq. (3); Section IV-C, Eqs. (13)-(18)] The pose estimation underlying every local frame relies solely on Levenberg-Marquardt minimization of SDF residuals, with the motion model set to 1 and no odometry fusion, loop closure, or temporal prior. The paper provides no error statistics for these pose estimates and no quantitative measure of inter-frame drift. Since the local frames, topological arcs g(LFi, LFj), and navigation all inherit pose errors from this registration, the central claim requires a direct evaluation of full-trajectory pose accuracy and drift, rather than only per-local-frame aggregates.
  3. [Section V-D, Table III] The navigation comparison is based on a single run with no description of the obstacle layout, number of trials, baseline tuning, or variance across runs. The claims that the approach is '5x faster and 50% shorter' than baseline RRT* are therefore anecdotal. The authors should provide repeated trials with statistical summaries (mean and standard deviation), a description of the environment and goal configurations, and ideally a comparison with a standard global planner under the same conditions.
minor comments (5)
  1. [Section V-A] 'Mean Root Square Error (MRSE)' should be 'Root Mean Square Error (RMSE),' and 'Absolute Trajectory Estimation (ATE)' should be 'Absolute Trajectory Error (ATE).'
  2. [Section II-C] There is a typo: 'Dynamic Window Approach (DWA) replies on velocity commands' should read 'relies on velocity commands.'
  3. [Section IV-B] The phrase 'We use strife sampling' appears to be a typo for 'stratified sampling' or similar; please clarify the intended sampling scheme.
  4. [Section V-C, Table II] The table header 'ATE RSME' contains a typo; it should read 'ATE RMSE.'
  5. [Section IV-C] The Jacobian in Eq. (17) and update in Eq. (18) are written in a compressed notation; adding explicit definitions of the rotation and translation parametrization in the tangent space would improve reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: central claims are benchmarked against external TUM ground-truth poses and a standard RRT* baseline, and the self-citations are auxiliary rather than load-bearing.

full rationale

The paper's load-bearing claims are evaluated against external benchmarks: Table II compares local-frame ATE RMSE against iMAP, NICE-SLAM, and ESLAM using TUM ground-truth poses, and Table III compares runtime and travel distance against a standard RRT* baseline. The pose estimation pipeline (Eqs. 13-18) solves an observation-to-model registration by minimizing SDF residuals; this is a standard joint estimation procedure, not a quantity defined in terms of the reported error. The self-cited works [8] and [27] appear only in a list of semantic segmentation sources alongside independent alternatives (YOLOv12 [25], SAM [26]) and are not used to justify any uniqueness or derivation. The paper does concede 'some drift between the local frames' and labels its metric an 'approximate RMSE,' which is a soundness and commensurability concern about whether Table II supports the headline precision claim, but it is not circularity: the reported numbers are not forced by construction. No derivation step reduces to its own input, and no fitted parameter is renamed as a prediction.

Assumptions & free parameters 5 free parameters · 4 assumptions · 2 invented entities

The central claim rests mostly on domain assumptions and hand-chosen parameters inherited from the implicit-SLAM literature. The most load-bearing ad hoc axiom is setting the motion model to 1 (Section IV-A), which pushes all pose estimation onto SDF registration. Several load-bearing parameters (local-frame creation thresholds, navigation angle alpha, sampling radius r and limit lb) are not quantified in the text, weakening reproducibility. No new physical entities are introduced; the invented entities are representational constructs with no external falsifiable handle.

free parameters (5)
  • Adaptive training loop convergence threshold (thresl) = 0.0001
    Hand-chosen stopping criterion for SDF network training (Table I); affects model quality and training runtime.
  • Local frame creation thresholds (viewpoint and translation variance) = not specified
    Section IV-D creates a new local frame when accumulated viewpoint/translation variance exceeds thresholds; the values are never given, yet they determine the number and size of local frames and therefore the whole topological map.
  • Navigation alternative-ray angle (alpha) = not specified
    Section IV-E generates alternative direction rays within a miniature angle alpha of the goal vector; the value controls how the planner escapes local obstacles and is not reported.
  • Surface sampling parameters (front/behind ratios, radius r, limit lb) = l in (0.3, 0.99) and (1.01, 1+lb)
    Section IV-B sets hand-chosen ranges for in/out-of-distribution training data generation; the on-surface radius r and behind-surface limit lb are not quantified.
  • Loss weights and learning rate (lambda_b, lambda_e, lambda_ewc, lambda) = 1.0, 0.5, 0.1, 0.001
    Hand-chosen training hyperparameters (Table I) standard in the cited PIN-SLAM/EWC framework.
assumptions (4)
  • ad hoc to paper Motion model P(xi_{t+1} | u_t, xi_t) is set to 1 (Section IV-A, before Eq. 3)
    The paper drops the pose transition term from the SLAM formulation, asserting the next pose is predictable from the previous pose and current velocity. This removes all motion priors and makes the perceptual model the sole source of pose information.
  • domain assumption The surface is perpendicular to the ray from camera to surface point when sampling on-surface training points (Section IV-B)
    Training data generation draws on-surface samples within a small radius on a plane built from the camera-to-point normal; this is approximate for oblique views and affects learned SDF quality.
  • domain assumption Once aligned, point clouds can be treated as complete surfaces with true SDF equal to zero (Section IV-C)
    Registration assumes each observation lies on the modeled surface after alignment; occlusions and incomplete views violate this.
  • domain assumption The perceptual model is approximated by the sigmoid of a neural SDF, phi(F_theta) approximately 1 - P(o_t | m, xi_t) (Section IV-B, Eq. 7)
    The probability that a colored point set belongs to the observed surface is modeled by a learned signed distance function; this is the standard reduction used in implicit SLAM and is stated without proof.
invented entities (2)
  • Spatial-implicit local frame
    purpose: A local map unit storing 3D points with color, learned features, and semantic labels; used for robot localization and obstacle avoidance, and serving as a node of the global topological map.
    The paper's main representational construct. It is validated only by the paper's own experiments (TUM local frames and in-lab runs); as a software representation it has no external falsifiable handle such as an independent predicted measurement.
  • Hybrid spatial-implicit point (p_j, f_j, s_j)
    purpose: Sparse key point combining 3D coordinates, interpolated learned features, and a semantic label (Eq. 5); used for frame similarity checks and as the atomic map element.
    Same representational construct as the local frame; no independent evidence outside the paper's experiments, so independent_evidence is False.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Bio-Inspired Hybrid Map: Spatial Implicit Local Frames and Topological Map for Mobile Cobot Navigation." pith.science (2026). https://pith.science/paper/MCBL5FT6

@misc{pith2026250704649,
  author       = {Pith},
  title        = {Pith review of: Bio-Inspired Hybrid Map: Spatial Implicit Local Frames and Topological Map for Mobile Cobot Navigation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MCBL5FT6}},
  note         = {Machine review of arXiv:2507.04649}
}
read the original abstract

Navigation is a fundamental capacity for mobile robots, enabling them to operate autonomously in complex and dynamic environments. Conventional approaches use probabilistic models to localize robots and build maps simultaneously using sensor observations. Recent approaches employ human-inspired learning, such as imitation and reinforcement learning, to navigate robots more effectively. However, these methods suffer from high computational costs, global map inconsistency, and poor generalization to unseen environments. This paper presents a novel method inspired by how humans perceive and navigate themselves effectively in novel environments. Specifically, we first build local frames that mimic how humans represent essential spatial information in the short term. Points in local frames are hybrid representations, including spatial information and learned features, so-called spatial-implicit local frames. Then, we integrate spatial-implicit local frames into the global topological map represented as a factor graph. Lastly, we developed a novel navigation algorithm based on Rapid-Exploring Random Tree Star (RRT*) that leverages spatial-implicit local frames and the topological map to navigate effectively in environments. To validate our approach, we conduct extensive experiments in real-world datasets and in-lab environments. We open our source code at https://github.com/tuantdang/simn}{https://github.com/tuantdang/simn.

Figures

Figures reproduced from arXiv: 2507.04649 by the authors.

Figure 1
Figure 1. Bio-inspired hybrid map concept: the robot aims to con￾tinuously learn and construct precise 3D semantic local frames using spatial information and learned features for effective collision avoidance while creating a topological map connecting local frames for global guidance, facilitating navigation tasks. Map building heavily depends on sensor quality and precise calibration, making it challenging to create consist… view at source ↗
Figure 2
Figure 2. System Overview In the proposed system, we begin by constructing sparse local maps using map-specific implicit spatial features. To do this, we begin by sampling 3D points from a sparse representation of the surface obtained from robot sensors to create a denser and more diverse training representation. This process entails sampling points along rays originating from both the front and back surfaces and selecting po… view at source ↗
Figure 3
Figure 3. (a) In-distribution and out-distribution data generation. (b) Plane-radius sampling for points on the surface . We define a map as a set of 3D points with the following learning features: m = {mj = (pj , fj , sj ) | pj ∈ R 3 , fj ∈ R m, sj ∈ N} (5) Here, fj represents the per-point learned features gener￾ated through interpolation with the features of neighboring points, assigning more weight to the points closest t… view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Building a 3D local frame from real-world office environ￾ments with semantic meaning: (a1 − a2) RGB images, (b1 − b2) depth images, (c1 − c2) identified traversable region on the floor (red polygons), (d) integration into the 3D local frame with the traversable region …
Figure 5
Figure 5. Figure 5: Qualitative results of constructing local frames and topological map from multiple observations: (a) variance in viewpoints, (b) variance in translation, (c) variance in viewpoints and translation, (d) finding the exit. same method for the fr2/xyz and fr3/office sequen…
Figure 6
Figure 6. Figure 6: Left: conventional RRT* approach. Right: Our approach with local frames and topological map. Whenever the robot moves, it must adjust its navigation plan to adapt to the local environment with respect to the topological map, as shown in [PITH_FULL_IMAGE:figures/full_f…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 20 canonical work pages

  1. [1]

    Orb-slam2: An open-source slam system for monocular, stereo, and rgb-d cameras,

    R. Mur-Artal and J. D. Tard ´os, “Orb-slam2: An open-source slam system for monocular, stereo, and rgb-d cameras,” IEEE transactions on robotics, vol. 33, no. 5, pp. 1255–1262, 2017

    2017

  2. [2]

    Dtam: Dense tracking and mapping in real-time,

    R. A. Newcombe, S. J. Lovegrove, and A. J. Davison, “Dtam: Dense tracking and mapping in real-time,” in 2011 international conference on computer vision . IEEE, 2011, pp. 2320–2327

    2011

  3. [3]

    Bad slam: Bundle adjusted direct rgb-d slam,

    T. Schops, T. Sattler, and M. Pollefeys, “Bad slam: Bundle adjusted direct rgb-d slam,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2019, pp. 134–144

    work page 2019

  4. [4]

    Elasticfusion: Dense slam without a pose graph

    T. Whelan, S. Leutenegger, R. F. Salas-Moreno, B. Glocker, and A. J. Davison, “Elasticfusion: Dense slam without a pose graph.” in Robotics: science and systems , vol. 11. Rome, Italy, 2015, p. 3

    2015

  5. [5]

    Codeslam—learning a compact, optimisable representation for dense visual slam,

    M. Bloesch, J. Czarnowski, R. Clark, S. Leutenegger, and A. J. Davison, “Codeslam—learning a compact, optimisable representation for dense visual slam,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2018, pp. 2560–2568

    work page 2018

  6. [6]

    Orbeez-slam: A real-time monocular visual slam with orb features and nerf-realized mapping,

    C.-M. Chung, Y .-C. Tseng, Y .-C. Hsu, X.-Q. Shi, Y .-H. Hua, J.-F. Yeh, W.-C. Chen, Y .-T. Chen, and W. H. Hsu, “Orbeez-slam: A real-time monocular visual slam with orb features and nerf-realized mapping,” in 2023 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2023, pp. 9400–9406

    work page 2023

  7. [7]

    Deepfactors: Real-time probabilistic dense monocular slam,

    J. Czarnowski, T. Laidlow, R. Clark, and A. J. Davison, “Deepfactors: Real-time probabilistic dense monocular slam,” IEEE Robotics and Automation Letters, vol. 5, no. 2, pp. 721–728, 2020

    2020

  8. [8]

    V3d-slam: Robust rgb-d slam in dynamic environments with 3d semantic geometry voting,

    T. Dang, K. Nguyen, and M. Huber, “V3d-slam: Robust rgb-d slam in dynamic environments with 3d semantic geometry voting,” in 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2024, pp. 7847–7853

    work page 2024

Show all 30 references
  1. [9]

    Se- manticfusion: Dense 3d semantic mapping with convolutional neural networks,

    J. McCormac, A. Handa, A. Davison, and S. Leutenegger, “Se- manticfusion: Dense 3d semantic mapping with convolutional neural networks,” in 2017 IEEE International Conference on Robotics and automation (ICRA). IEEE, 2017, pp. 4628–4635

    2017

  2. [10]

    Nodeslam: Neural object descriptors for multi-view shape reconstruction,

    E. Sucar, K. Wada, and A. Davison, “Nodeslam: Neural object descriptors for multi-view shape reconstruction,” in 2020 International Conference on 3D Vision (3DV) . IEEE, 2020, pp. 949–958

    2020

  3. [11]

    Scenecode: Monocular dense semantic reconstruction using learned encoded scene representations,

    S. Zhi, M. Bloesch, S. Leutenegger, and A. J. Davison, “Scenecode: Monocular dense semantic reconstruction using learned encoded scene representations,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2019, pp. 11 776–11 785

    2019

  4. [12]

    imap: Implicit map- ping and positioning in real-time,

    E. Sucar, S. Liu, J. Ortiz, and A. J. Davison, “imap: Implicit map- ping and positioning in real-time,” in Proceedings of the IEEE/CVF international conference on computer vision , 2021, pp. 6229–6238

    2021

  5. [13]

    Eslam: Efficient dense slam system based on hybrid representation of signed distance fields,

    M. M. Johari, C. Carta, and F. Fleuret, “Eslam: Efficient dense slam system based on hybrid representation of signed distance fields,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 17 408–17 419

    2023

  6. [14]

    Trajectory encoding in the hippocampus and entorhinal cortex,

    L. M. Frank, E. N. Brown, and M. Wilson, “Trajectory encoding in the hippocampus and entorhinal cortex,” Neuron, vol. 27, no. 1, pp. 169–178, 2000

    2000

  7. [15]

    Memory, navigation and theta rhythm in the hippocampal-entorhinal system,

    G. Buzs ´aki and E. I. Moser, “Memory, navigation and theta rhythm in the hippocampal-entorhinal system,” Nature neuroscience, vol. 16, no. 2, pp. 130–138, 2013

    2013

  8. [16]

    Viewpoints: how the hippocampus contributes to memory, navigation and cognition,

    J. Lisman, G. Buzs ´aki, H. Eichenbaum, L. Nadel, C. Ranganath, and A. D. Redish, “Viewpoints: how the hippocampus contributes to memory, navigation and cognition,” Nature neuroscience, vol. 20, no. 11, pp. 1434–1447, 2017

    2017

  9. [17]

    Robust monte carlo localization for mobile robots,

    S. Thrun, D. Fox, W. Burgard, and F. Dellaert, “Robust monte carlo localization for mobile robots,” Artificial intelligence , vol. 128, no. 1-2, pp. 99–141, 2001

    2001

  10. [18]

    Learning Model Predictive Controllers with Real-Time Attention for Real-World Navigation,

    X. Xiao, T. Zhang, K. Choromanski, E. Lee, A. Francis, J. Varley, S. Tu, S. Singh, P. Xu, F. Xia, S. M. Persson, D. Kalashnikov, L. Takayama, R. Frostig, J. Tan, C. Parada, and V . Sindhwani, “Learning Model Predictive Controllers with Real-Time Attention for Real-World Naviga...

    2022 arXiv

  11. [19]

    Predictive collision avoidance for the dynamic window approach,

    M. Missura and M. Bennewitz, “Predictive collision avoidance for the dynamic window approach,” in 2019 International Conference on Robotics and Automation (ICRA) . IEEE, 2019, pp. 8620–8626

    2019

  12. [20]

    Recent advances in rapidly-exploring random tree: A review,

    T. Xu, “Recent advances in rapidly-exploring random tree: A review,” Heliyon, 2024

    2024

  13. [21]

    Integrating topological and metric maps for mobile robot navigation: A statistical approach,

    S. Thrun, J.-S. Gutmann, D. Fox, W. Burgard, B. Kuipers, et al. , “Integrating topological and metric maps for mobile robot navigation: A statistical approach,” AAAI/IAAI, vol. 9, pp. 989–995, 1998

    1998

  14. [22]

    DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation,

    J. J. Park, P. Florence, J. Straub, R. Newcombe, and S. Lovegrove, “DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation,” 2019

    2019

  15. [23]

    PIN-SLAM: LiDAR SLAM Using a Point-Based Implicit Neu- ral Representation for Achieving Global Map Consistency,

    Y . Pan, X. Zhong, L. Wiesmann, T. Posewsky, J. Behley, and C. Stach- niss, “PIN-SLAM: LiDAR SLAM Using a Point-Based Implicit Neu- ral Representation for Achieving Global Map Consistency,” IEEE Transactions on Robotics , vol. 40, 2024, conference Name: IEEE Transactions on Robotics

    2024

  16. [24]

    Overcoming catastrophic forgetting in neural networks,

    J. Kirkpatrick, R. Pascanu, N. Rabinowitz, J. Veness, G. Desjardins, A. A. Rusu, K. Milan, J. Quan, T. Ramalho, A. Grabska-Barwinska, et al., “Overcoming catastrophic forgetting in neural networks,” Pro- ceedings of the national academy of sciences , vol. 114, no. 13, pp. 3521...

    2017

  17. [25]

    Yolov12: Attention-centric real- time object detectors,

    Y . Tian, Q. Ye, and D. Doermann, “Yolov12: Attention-centric real- time object detectors,” arXiv preprint arXiv:2502.12524 , 2025

    2025 arXiv

  18. [26]

    Segment anything,

    A. Kirillov, E. Mintun, N. Ravi, H. Mao, C. Rolland, L. Gustafson, T. Xiao, S. Whitehead, A. C. Berg, W.-Y . Lo, et al. , “Segment anything,” in Proceedings of the IEEE/CVF international conference on computer vision , 2023, pp. 4015–4026

    2023

  19. [27]

    V olumetric mapping with panoptic refinement via kernel density estimation for mobile robots,

    K. Nguyen, T. Dang, and M. Huber, “V olumetric mapping with panoptic refinement via kernel density estimation for mobile robots,” arXiv preprint arXiv:2412.11241 , 2024

    2024 arXiv

  20. [28]

    NICE-SLAM: Neural Implicit Scalable Encoding for SLAM,

    Z. Zhu, S. Peng, V . Larsson, W. Xu, H. Bao, Z. Cui, M. R. Oswald, and M. Pollefeys, “NICE-SLAM: Neural Implicit Scalable Encoding for SLAM,” in 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . New Orleans, LA, USA: IEEE, June 2022, pp. 12 776–12 78...

    2022

  21. [29]

    ESLAM: Efficient Dense SLAM System Based on Hybrid Representation of Signed Distance Fields,

    M. M. Johari, C. Carta, and F. Fleuret, “ESLAM: Efficient Dense SLAM System Based on Hybrid Representation of Signed Distance Fields,” Apr. 2023, arXiv:2211.11704 [cs]. [Online]. Available: http://arxiv.org/abs/2211.11704

    2023 arXiv

  22. [30]

    A benchmark for the evaluation of rgb-d slam systems,

    J. Sturm, N. Engelhard, F. Endres, W. Burgard, and D. Cremers, “A benchmark for the evaluation of rgb-d slam systems,” Proc. of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pp. 573–580, 2012. [Online]. Available: https://vision.in.tum.de/dat...

    2012

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.