Pith. sign in

REVIEW 4 major objections 6 minor 38 references

What Really Matters for Robust Multi-Sensor HD Map Construction?

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Adding augmentation, cross-modal attention, and modality dropout makes camera-LiDAR HD map construction more accurate and far more resilient to sensor failure.

desk verdict The clean nuScenes SOTA is real and useful; the 14.3-point robustness gain is not yet supported. read the letter →

arxiv 2507.01484 v1 pith:RIKWUEWZ submitted 2025-07-02 cs.CV

Xiaoshuai Hao , Yuting Zhao , Yuheng Ji , Luanyuan Dai , Peng Hao , Dingzhe Li , Shuai Cheng , Rong Yin This is my paper
classification cs.CV
keywords HDmapconstructioncamera-LiDARfusionmulti-sensorrobustnesssensorcorruptionmodalitydropoutcross-modaltransformerBEVautonomousdriving
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 tries to establish that the frailty of camera-LiDAR HD map construction under sensor corruption is fixable without giving up accuracy. It argues that three training-time ingredients—camera and LiDAR data augmentation, a cross-modal interaction transformer that lets each sensor's BEV features borrow context from the other, and a modality dropout that randomly disables one sensor during training—make existing fusion models markedly more robust to 13 synthetic corruption types. On the nuScenes validation set the full recipe lifts the HIMap baseline from 74.3 to 77.0 mAP and raises its mean Resilience Score by 14.3 points, with comparable gains on MapTR. A sympathetic reader would care because autonomous vehicles need map models that keep working when sensors fail, not just on clean benchmarks.

What carries the argument

The mechanism that carries the argument is the package of three training-time components, with the Cross-Modal Interaction Transformer (CIT) as the novel fusion piece. CIT flattens camera and LiDAR BEV features into token sequences $T^{BEV}_{Camera}, T^{BEV}_{LiDAR} \in \mathbb{R}^{HW \times C}$, concatenates them with a learnable positional embedding, applies multi-head self-attention $Z=\mathrm{softmax}(QK^T/\sqrt{D_k})V$ with a residual MLP, and then splits the refined tokens back into per-modality features that enter a dynamic fusion module. Around it sit GridMask image augmentation and LiDAR point dropout, plus a modality dropout training rule that with probability $p_{md}$ drops one modality, keeping LiDAR with probability $p_L$ and camera with probability $1-p_L$; this rule is what forces the network to stay functional when a sensor disappears at test time.

What would settle it

A decisive test would be to evaluate the trained RoboMap and its baselines on a corruption distribution not used to build the Multi-Sensor Corruption dataset—for instance, real fog and snow footage, hardware-induced LiDAR misalignment, or randomly dropped camera frames at deployment time. If the 14.3-point mRS advantage over HIMap does not reproduce under those held-out conditions, or if the clean-data mAP gain reverses, the robustness claim is confined to the synthetic benchmark.

Watch

Extended reading notes

Core claim

The paper's central claim is that robustness in multi-modal HD map construction comes from combining complementary training-time strategies, not from any single module. To test this, the paper constructs a Multi-Sensor Corruption dataset of 13 corruption types at three severity levels from the nuScenes validation set, covering camera-only, LiDAR-only, and combined failures. Individually, the fusion module, data augmentation, and dropout training each improve clean accuracy but leave distinct resilience weak spots; together they raise mean Resilience Score from 67.69 to 81.99 for HIMap and from 69.51 to 79.06 for MapTR, while setting a new state of the art of 77.0 mAP on clean nuScenes. The claim is that these gains reflect the model learning to lean on whichever modality remains intact.

Load-bearing premise

The load-bearing premise is that the 13 synthetic corruption types on the nuScenes validation set, at the three severity levels, faithfully represent the sensor failures and weather effects a real autonomous vehicle encounters; if real-world corruptions differ from these synthetic ones, the measured resilience gains may not transfer to deployment.

Editorial extensions

If this is right

  • Existing camera-LiDAR fusion HD map models can be made more robust by adding the three components without architectural redesign, since they are plug-and-play and improve both MapTR and HIMap.
  • Under complete single-sensor failure, such as a camera crash or lost frames, the model retains far more mapping accuracy than the same model trained without modality dropout.
  • The clean-data gains, such as 77.0 mAP for the HIMap-based RoboMap, mean robustness training does not have to be traded against benchmark accuracy.
  • The mean Resilience Score becomes a usable comparison metric: future fusion methods can report mRS and mRRS against these baselines on the same 13 corruption types.

Reading between the lines

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

  • The paper does not tune the dropout probabilities $p_{md}$ and $p_L$; a natural follow-up is to adjust them per corruption family, since single-sensor failures such as camera crash and frame lost gain most from aggressive dropout while temporal misalignment gains least.
  • The recipe is not specific to HD map construction, so the same three components could plausibly harden other camera-LiDAR BEV perception tasks such as 3D object detection and BEV segmentation; the paper evaluates only map construction.
  • CIT's self-attention runs over roughly $2HW$ BEV tokens, so fusion cost scales quadratically with BEV resolution; a windowed or linear-attention variant would be the natural next step for high-resolution maps, a scalability question the paper leaves open.
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

4 major / 6 minor

Summary. The paper proposes RoboMap, a framework that combines camera and LiDAR data augmentation, a Cross-Modal Interaction Transformer (CIT) fusion module, and a modality dropout training strategy to improve the robustness of multi-modal HD map construction under sensor corruptions. The authors construct a Multi-Sensor Corruption dataset with 13 corruption types on the nuScenes validation set and define Resilience Score (RS) and Relative Resilience Score (RRS) metrics. They report that RoboMap improves the mean Resilience Score over MapTR by 9.55 and over HIMap by 14.3, while also improving clean-validation mAP to 77.0 (from HIMap's 74.3). The clean-data mAP comparison is grounded in the external nuScenes benchmark, but the robustness claims rest entirely on an author-built, unreleased corruption benchmark with unspecified generation parameters and no error bars, and the named Dynamic Fusion module is not technically defined.

Significance. If the robustness gains are real and reproducible, the paper would offer a simple, plug-and-play recipe for making camera–LiDAR HD map construction resilient to sensor failures without sacrificing clean-data accuracy. The clean-data result is a direct measurement against an external benchmark, which gives that part of the claim independent grounding. The ablation structure is clear and the three components are individually motived. However, the central robustness contribution is measured only on a self-created benchmark that is not released, not cross-validated against existing robustness corpora, and lacks any uncertainty quantification. The undefined 'Dynamic Fusion' module further prevents independent verification of the fusion component. These issues are load-bearing for the paper's central claim and must be resolved before the robustness conclusions can be accepted.

major comments (4)
  1. [Section III] The Multi-Sensor Corruption dataset is the sole basis for the robustness claims (mRS gains in Tables II and III), but it is not released and its construction is under-specified. The paper lists 13 corruption types but does not provide the generation algorithms, parameter settings, or severity levels for the Easy/Moderate/Hard variants, deferring instead to the authors' own protocols in [19] and [20]. No standard deviation across random seeds is reported. Consequently, the 14.3-point mRS improvement over HIMap cannot be independently reproduced, and the benchmark cannot be compared against other robustness suites. Please release the dataset, specify the exact corruption parameters for each severity level, and report mean ± std over at least three training seeds.
  2. [Section IV] The 'Dynamic Fusion' module is invoked but never defined. After Eq. (5), the text states 'We utilize the Dynamic Fusion module to aggregate the multi-modal BEV feature inputs, ˆF BEV Camera and ˆF BEV LiDAR, resulting in the aggregated features Ffused,' but no architecture, equations, or hyperparameters are provided for this module. Without a concrete definition, the fusion component cannot be independently implemented or tested, and the attribution of robustness gains to the fusion module is not verifiable. Please provide a formal description of the Dynamic Fusion operation, or replace it with a concretely specified aggregation mechanism.
  3. [Section V.D, Tables II and III] The robustness evaluation contains no comparison with existing robustness methods or benchmarks. The authors compare only their own baselines and variants. To establish that the observed mRS gains are not simply an artifact of fitting the authors' specific corruption suite, the paper should compare against established robustness protocols (e.g., MapBench [19], MSC-Bench [20], and RoboBEV [14]) and report the RRS values with confidence intervals. Without such comparisons, the claim that the proposed strategies 'significantly enhance the robustness of multi-modal fusion methods' remains unsubstantiated beyond the internal benchmark.
  4. [Section I and Section II.A] The claim that 'RoboMap is the first study to systematically explore the robustness of HD map construction under multi-sensor corruptions' is contradicted by the authors' own references [19] (MapBench) and [20] (MSC-Bench), both of which benchmark HD map construction under sensor corruptions. The novelty statement should be reformulated to clearly delineate what RoboMap adds beyond those prior benchmarks, rather than asserting firstness that the related work already rebuts.
minor comments (6)
  1. [Abstract and Section V.A] The abstract states that the method is 'evaluated on a challenging dataset containing 10 days of NuScenes data,' while Section V.A says the nuScenes validation set is used. The validation set contains 150 scenes, not '10 days'; please make the description consistent.
  2. [Equation (1)] In the definition of RS, the denominator uses Acc_clean, but it is not clarified whether Acc_clean is the clean mAP of the same model or a fixed reference value. This ambiguity affects the interpretation of the RRS metric and should be stated explicitly.
  3. [Section IV, paragraph after Eq. (5)] The notation 'Ffused' and 'Ff used' appears inconsistent; please unify the notation for the fused feature.
  4. [References] References [26] and [31] are the same paper (Zhang et al., 'Online vectorized HD map construction using geometry'); please deduplicate.
  5. [Figure 2 caption] The caption compresses the architecture description and does not explain the 'Dynamic Fusion' box; please expand the caption so the figure is self-contained.
  6. [General] The paper does not state whether code or the Multi-Sensor Corruption dataset will be released. Given that the robustness benchmark is central to the claims, a clear data/code availability statement is needed.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: clean-data SOTA is externally benchmarked and the robustness gains are empirical measurements, not reductions to the paper's own definitions or fitted inputs.

full rationale

No step in the claimed derivation chain makes a prediction equivalent to an input by construction. The clean mAP result in Table I is benchmarked against external nuScenes validation numbers for published methods such as MapTR and HIMap, so the accuracy claim has independent grounding. The robustness claim is an empirical comparison of mean Resilience Scores defined in Eq. (1), computed from measured accuracies on the author-constructed Multi-Sensor Corruption dataset; the metric is an arithmetic definition and does not encode the outcome. Although the corruption protocol is inherited from the authors' own MapBench [19] and MSC-Bench [20] and is not fully specified or released, this is a reproducibility and external-validity limitation, not circularity. Similarly, using GridMask, LiDAR dropout, and modality dropout during training to match the failure modes in the test suite is a domain-matched training choice; it may weaken generalization claims, but it does not make the evaluation result equivalent to the training recipe by definition. The undefined 'Dynamic Fusion' module is a missing implementation detail, not a circular step. No uniqueness theorem or load-bearing result is imported from self-citations to force the paper's choices.

Assumptions & free parameters 3 free parameters · 3 assumptions · 1 invented entities

The central claim rests on a small number of hand-chosen hyperparameters (modality dropout probabilities) that are not reported, on the acceptance of the authors' own corruption protocol as a valid proxy for real-world failures, and on an unspecified Dynamic Fusion module. These are the load-bearing choices that a reader must accept to trust the robustness results.

free parameters (3)
  • modality dropout probability pmd = not reported
    Probability of randomly dropping either the camera or LiDAR BEV features during training; central to the modality dropout strategy but its value is never given.
  • LiDAR retention probability pL = not reported
    Probability of keeping LiDAR when a modality is dropped (camera retention is 1-pL); value not stated.
  • GridMask and point-dropout intensities = not reported
    Strengths of image masking and LiDAR point removal are not specified, yet the paper credits these augmentations with robustness gains.
assumptions (3)
  • domain assumption The corruption synthesis protocol of [19] and [20] is a valid model of real-world sensor failures.
    Section III says the dataset is 'following the protocol established in [19], [20]' but no independent evidence links the synthetic corruptions to real sensor degradation.
  • standard math Standard transformer self-attention (Q, K, V) is correct.
    Equations (3)-(5) invoke standard multi-head attention.
  • ad hoc to paper The Dynamic Fusion module aggregates features as claimed.
    Section IV references a 'Dynamic Fusion' module without describing it, so the fused-feature claim rests on an unexamined component.
invented entities (1)
  • Dynamic Fusion module
    purpose: Aggregates camera and LiDAR BEV features after the Cross-Modal Interaction Transformer.
    Mentioned by name in Section IV but never defined with equations, architectural details, or a citation; it functions as an unexplained black box supporting the central result.

how reviews work

0 comments
Cite this review

Pith. "Pith review of What Really Matters for Robust Multi-Sensor HD Map Construction?." pith.science (2026). https://pith.science/paper/RIKWUEWZ

@misc{pith2026250701484,
  author       = {Pith},
  title        = {Pith review of: What Really Matters for Robust Multi-Sensor HD Map Construction?},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RIKWUEWZ}},
  note         = {Machine review of arXiv:2507.01484}
}
read the original abstract

High-definition (HD) map construction methods are crucial for providing precise and comprehensive static environmental information, which is essential for autonomous driving systems. While Camera-LiDAR fusion techniques have shown promising results by integrating data from both modalities, existing approaches primarily focus on improving model accuracy and often neglect the robustness of perception models, which is a critical aspect for real-world applications. In this paper, we explore strategies to enhance the robustness of multi-modal fusion methods for HD map construction while maintaining high accuracy. We propose three key components: data augmentation, a novel multi-modal fusion module, and a modality dropout training strategy. These components are evaluated on a challenging dataset containing 10 days of NuScenes data. Our experimental results demonstrate that our proposed methods significantly enhance the robustness of baseline methods. Furthermore, our approach achieves state-of-the-art performance on the clean validation set of the NuScenes dataset. Our findings provide valuable insights for developing more robust and reliable HD map construction models, advancing their applicability in real-world autonomous driving scenarios. Project website: https://robomap-123.github.io.

Figures

Figures reproduced from arXiv: 2507.01484 by the authors.

Figure 1
Figure 1. Overview of the Multi-Sensor Corruption dataset. Multi-Sensor Corruption includes 13 types of synthetic camera-LiDAR corruption combinations that perturb both camera and LiDAR inputs, either separately or concurrently. elements, such as lane boundaries and road structures. These maps are essential for accurate localization and path planning in autonomous driving systems. Recent advancements in camera-LiDAR fusion me… view at source ↗
Figure 2
Figure 2. Overview of the RoboMap Framework. The RoboMap framework begins by applying data augmentation to both camera images and LiDAR point clouds. Next, features are efficiently extracted from the multi-modal sensor inputs and transformed into a unified Bird’s-Eye View (BEV) space using view transformation techniques. We then introduce a novel multi-modal BEV fusion module to effectively integrate features from both modali… view at source ↗
Figure 3
Figure 3. Analyze the impact of different modules on the HD map construction [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Relative robustness visualization. Relative Resilience Score (RRS) computed with mAP using original MapTR [5] as baseline. [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Relative robustness visualization. Relative Resilience Score (RRS) computed with mAP using original HIMap [29] as baseline. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

38 extracted references · 27 canonical work pages

  1. [19]

    Is your hd map constructor reliable under sensor corruptions?

    X. Hao, M. Wei, Y . Yang, H. Zhao, H. Zhang, Y . Zhou, Q. Wang, W. Li, L. Kong, and J. Zhang, “Is your hd map constructor reliable under sensor corruptions?” in Advances in Neural Information Processing Systems, vol. 37, 2024

    work page 2024

  2. [20]

    Msc-bench: Benchmarking and analyzing multi-sensor corruption for driving perception,

    X. Hao, G. Liu, Y . Zhao, Y . Ji, M. Wei, H. Zhao, L. Kong, R. Yin, and Y . Liu, “Msc-bench: Benchmarking and analyzing multi-sensor corruption for driving perception,” arXiv preprint arXiv:2501.01037 , 2025

    arXiv 2025

  3. [14]

    Robobev: Towards robust bird’s eye view perception under corruptions,

    S. Xie, L. Kong, W. Zhang, J. Ren, L. Pan, K. Chen, and Z. Liu, “Robobev: Towards robust bird’s eye view perception under corruptions,” arXiv preprint arXiv:2304.06719 , 2023

    arXiv 2023

  4. [1]

    Mapdistill: Boosting efficient camera-based hd map construction via camera-lidar fusion model distillation,

    X. Hao, R. Li, H. Zhang, D. Li, R. Yin, S. Jung, S.-I. Park, B. Yoo, H. Zhao, and J. Zhang, “Mapdistill: Boosting efficient camera-based hd map construction via camera-lidar fusion model distillation,” in European Conference on Computer Vision , 2024, pp. 166–183

    work page 2024

  5. [2]

    Stream query denoising for vectorized hd-map construction,

    S. Wang, F. Jia, W. Mao, Y . Liu, Y . Zhao, Z. Chen, T. Wang, C. Zhang, X. Zhang, and F. Zhao, “Stream query denoising for vectorized hd-map construction,” in European Conference on Computer Vision , 2024, pp. 203–220

    work page 2024

  6. [3]

    Stvit+: improving self-supervised multi-camera depth estimation with spatial-temporal context and adversarial geometry regularization,

    Z. Chen, H. Zhao, X. Hao, B. Yuan, and X. Li, “Stvit+: improving self-supervised multi-camera depth estimation with spatial-temporal context and adversarial geometry regularization,” Applied Intelligence , vol. 55, no. 5, p. 328, 2025

    work page 2025

  7. [4]

    Diffmap: Enhancing map segmentation with map prior using diffusion model,

    P. Jia, T. Wen, Z. Luo, M. Yang, K. Jiang, Z. Liu, X. Tang, Z. Lei, L. Cui, B. Zhang et al. , “Diffmap: Enhancing map segmentation with map prior using diffusion model,” IEEE Robotics and Automation Letters, 2024

    work page 2024

  8. [5]

    Maptr: Structured modeling and learning for online vectorized hd map construction,

    B. Liao, S. Chen, X. Wang, T. Cheng, Q. Zhang, W. Liu, and C. Huang, “Maptr: Structured modeling and learning for online vectorized hd map construction,” in International Conference on Learning Representations , 2023

    2023

Show all 38 references
  1. [6]

    Fastrsr: Efficient and accurate road surface reconstruction from bird’s eye view,

    Y . Zhao, Y . Ji, X. Hao, and S. Li, “Fastrsr: Efficient and accurate road surface reconstruction from bird’s eye view,” arXiv preprint arXiv:2504.09535, 2025

    2025 arXiv

  2. [7]

    Mapfusion: A novel bev feature fusion network for multi-modal map construction,

    X. Hao, Y . Diao, M. Wei, Y . Yang, P. Hao, R. Yin, H. Zhang, W. Li, S. Zhao, and Y . Liu, “Mapfusion: A novel bev feature fusion network for multi-modal map construction,” Information Fusion , vol. 119, p. 103018, 2025

    2025

  3. [8]

    Bevfusion: A simple and robust lidar-camera fusion framework,

    T. Liang, H. Xie, K. Yu, Z. Xia, Z. Lin, Y . Wang, T. Tang, B. Wang, and Z. Tang, “Bevfusion: A simple and robust lidar-camera fusion framework,” 2022, pp. 10 421–10 434

    2022

  4. [9]

    Deepfusionmot: A 3d multi- object tracking framework based on camera-lidar fusion with deep association,

    X. Wang, C. Fu, Z. Li, Y . Lai, and J. He, “Deepfusionmot: A 3d multi- object tracking framework based on camera-lidar fusion with deep association,” IEEE Robotics and Automation Letters , pp. 8260–8267, 2022

    2022

  5. [10]

    Safemap: Robust hd map construction from incomplete observations,

    X. Hao, L. Kong, R. Yin, P. Wang, J. Zhang, Y . Diao, and S. Zhao, “Safemap: Robust hd map construction from incomplete observations,” in F orty-second International Conference on Machine Learning , 2025

    2025

  6. [11]

    Team samsung-ral: Technical report for 2024 robodrive challenge- robust map segmentation track,

    X. Hao, Y . Yang, H. Zhang, M. Wei, Y . Zhou, H. Zhao, and J. Zhang, “Team samsung-ral: Technical report for 2024 robodrive challenge- robust map segmentation track,” arXiv preprint arXiv:2405.10567 , 2024

    2024 arXiv

  7. [12]

    The robodrive challenge: Drive anytime anywhere in any condition,

    L. Kong, S. Xie, H. Hu, Y . Niu, W. T. Ooi, B. R. Cottereau, L. X. Ng, Y . Ma, W. Zhang, L. Pan et al. , “The robodrive challenge: Drive anytime anywhere in any condition,” arXiv preprint arXiv:2405.08816 , 2024

    2024 arXiv

  8. [13]

    Using temporal information and mixing-based data augmentations for robust hd map construction

    X. Hao, Y . Yang, H. Zhang, M. Wei, Y . Zhou, H. Zhao, and J. Zhang, “Using temporal information and mixing-based data augmentations for robust hd map construction.”

  9. [15]

    Robustness-aware 3d object detection in autonomous driving: A review and outlook,

    Z. Song, L. Liu, F. Jia, Y . Luo, C. Jia, G. Zhang, L. Yang, and L. Wang, “Robustness-aware 3d object detection in autonomous driving: A review and outlook,” IEEE Transactions on Intelligent Transportation Systems , 2024

    2024

  10. [16]

    The four most basic elements in machine cognition,

    D. Li, J. Yin, T. Zhang, W. Han, and H. Bao, “The four most basic elements in machine cognition,” Data Intelligence , vol. 6, no. 2, pp. 297–319, 2024

    2024

  11. [17]

    Robo3d: Towards robust and reliable 3d perception against corruptions,

    L. Kong, Y . Liu, X. Li, R. Chen, W. Zhang, J. Ren, L. Pan, K. Chen, and Z. Liu, “Robo3d: Towards robust and reliable 3d perception against corruptions,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 19 994–20 006

    2023

  12. [18]

    Understanding the robustness of 3d object detection with bird’s-eye-view representations in autonomous driving,

    Z. Zhu, Y . Zhang, H. Chen, Y . Dong, S. Zhao, W. Ding, J. Zhong, and S. Zheng, “Understanding the robustness of 3d object detection with bird’s-eye-view representations in autonomous driving,” in Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognit...

    2023

  13. [21]

    nuscenes: A multimodal dataset for autonomous driving,

    H. Caesar, V . Bankiti et al. , “nuscenes: A multimodal dataset for autonomous driving,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2020, pp. 11 618–11 628

    2020

  14. [22]

    Hdmapnet: An online hd map construction and evaluation framework,

    Q. Li, Y . Wang, Y . Wang, and H. Zhao, “Hdmapnet: An online hd map construction and evaluation framework,” in International Conference on Robotics and Automation (ICRA) , 2022, pp. 4628–4634

    2022

  15. [23]

    Vectormapnet: End-to-end vectorized hd map learning,

    Y . Liu, T. Yuan, Y . Wang, Y . Wang, and H. Zhao, “Vectormapnet: End-to-end vectorized hd map learning,” in International Conference on Machine Learning , 2023, pp. 22 352–22 369

    2023

  16. [24]

    Pivotnet: Vectorized pivot learning for end-to-end hd map construction,

    W. Ding, L. Qiao, X. Qiu, and C. Zhang, “Pivotnet: Vectorized pivot learning for end-to-end hd map construction,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 3672–3682

    2023

  17. [25]

    End-to-end vectorized hd- map construction with piecewise bezier curve,

    L. Qiao, W. Ding, X. Qiu, and C. Zhang, “End-to-end vectorized hd- map construction with piecewise bezier curve,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 13 218–13 228

    2023

  18. [27]

    Maptrv2: An end-to-end framework for online vectorized hd map construction,

    B. Liao, S. Chen, Y . Zhang, B. Jiang, Q. Zhang, W. Liu, C. Huang, and X. Wang, “Maptrv2: An end-to-end framework for online vectorized hd map construction,” International Journal of Computer Vision , pp. 1–23, 2024

    2024

  19. [28]

    Streammapnet: Streaming mapping network for vectorized online hd map construction,

    T. Yuan, Y . Liu, Y . Wang, Y . Wang, and H. Zhao, “Streammapnet: Streaming mapping network for vectorized online hd map construction,” in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision , 2024, pp. 7356–7365

    2024

  20. [29]

    Himap: Hybrid representation learning for end-to-end vectorized hd map construction,

    Y . Zhou, H. Zhang, J. Yu, Y . Yang, S. Jung, S.-I. Park, and B. Yoo, “Himap: Hybrid representation learning for end-to-end vectorized hd map construction,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 15 396–15 406

    2024

  21. [30]

    Mbfusion: A new multi-modal bev feature fusion method for hd map construction,

    X. Hao, H. Zhang, Y . Yang, Y . Zhou, S. Jung, S.-I. Park, and B. Yoo, “Mbfusion: A new multi-modal bev feature fusion method for hd map construction,” in IEEE International Conference on Robotics and Automation (ICRA) , 2024, pp. 15 922–15 928

    2024

  22. [31]

    Online vectorized hd map construction using geometry,

    Z. Zhang, Y . Zhang, X. Ding, F. Jin, and X. Yue, “Online vectorized hd map construction using geometry,” in European Conference on Computer Vision, 2024, pp. 73–90

    2024

  23. [32]

    Mgmap: Mask-guided learning for online vectorized hd map construction,

    X. Liu, S. Wang, W. Li, R. Yang, J. Chen, and J. Zhu, “Mgmap: Mask-guided learning for online vectorized hd map construction,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 14 812–14 821

    2024

  24. [33]

    A survey on image data augmentation for deep learning,

    C. Shorten and T. M. Khoshgoftaar, “A survey on image data augmentation for deep learning,” Journal of big data , vol. 6, no. 1, pp. 1–48, 2019

    2019

  25. [34]

    Part-aware data augmentation for 3d object detection in point cloud,

    J. Choi, Y . Song, and N. Kwak, “Part-aware data augmentation for 3d object detection in point cloud,” in IEEE/RSJ International Conference on Intelligent Robots and Systems , 2021, pp. 3391–3397

    2021

  26. [35]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2016, pp. 770–778

    2016

  27. [36]

    Efficient and robust 2d-to-bev representation learning via geometry-guided kernel transformer,

    S. Chen, T. Cheng, X. Wang, W. Meng, Q. Zhang, and W. Liu, “Efficient and robust 2d-to-bev representation learning via geometry-guided kernel transformer,” arXiv preprint arXiv:2206.04584 , 2022

    2022 arXiv

  28. [37]

    SECOND: sparsely embedded convolutional detection,

    Y . Yan, Y . Mao, and B. Li, “SECOND: sparsely embedded convolutional detection,” Sensors, p. 3337, 2018

    2018

  29. [38]

    Bevfusion: Multi-task multi-sensor fusion with unified bird’s-eye view representation,

    Z. Liu, H. Tang, A. Amini, X. Yang, H. Mao, D. L. Rus, and S. Han, “Bevfusion: Multi-task multi-sensor fusion with unified bird’s-eye view representation,” in IEEE international conference on robotics and automation (ICRA) , 2023, pp. 2774–2781

    2023

  30. [39]

    Electronic device and method with birds-eye-view image processing,

    H. Xiaoshuai, Y . Zhou, B. I. Yoo, H. Zhang, S. Park, and S. Jung, “Electronic device and method with birds-eye-view image processing,” Mar. 13 2025, uS Patent App. 18/883,424

    2025

Pith tools

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