Pith. sign in

REVIEW 4 major objections 4 minor 51 references

LiDAR-Guided Monocular 3D Object Detection for Long-Range Railway Monitoring

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

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

Pith's one-line read A monocular camera, trained with LiDAR depth guidance, can detect and localize 3D railway objects up to 250 meters away.

desk verdict Sensible modular pipeline for monocular railway 3D detection, but the 250-meter claim is not backed by any end-to-end 3D metric; the authors are honest about the gap, making this a workshop-grade paper that needs either a full-pipeline 3D evaluation or a scaled-back claim. read the letter →

arxiv 2504.18203 v1 pith:RSLPB6SN submitted 2025-04-25 cs.CV cs.LG

Raul David Dominguez Sanchez , Xavier Diaz Ortiz , Xingcheng Zhou , Max Peter Ronecker , Michael Karner , Daniel Watzenig , Alois Knoll This is my paper · ORCID
classification cs.CVcs.LG
keywords monocular3Dobjectdetectionrailwayperceptiondepthestimationfrustum-basedLiDAR-guidedtraininglong-rangeOSDaR23autonomoustrains
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 claims that a single camera, trained with LiDAR depth supervision, can perform 3D object detection in railway scenes at distances up to 250 meters. The proposed pipeline, MFF, replaces the LiDAR point cloud used by the Faraway-Frustum approach with pseudo-clouds generated from learned monocular depth, so that no LiDAR is needed at test time. Evaluations on the OSDaR23 dataset show detection of persons, road vehicles, buffer stops, catenary poles, and signal poles at long range, and the long-range head outperforms Faraway-Frustum's head on KITTI car detection (60.61 vs. 46.90 3D mAP@0.1). This matters because trains require braking distances far exceeding those of cars, so extending perception beyond the typical automotive range is safety-critical.

What carries the argument

The load-bearing mechanism is the pseudo-cloud frustum: monocular depth maps are back-projected into 3D pseudo point clouds, and 2.5D object detections carve out pyramid-shaped frustums that are routed to dedicated short- and long-range detection heads. This replaces the LiDAR point cloud that the original Faraway-Frustum needed while preserving the frustum's role of narrowing the 3D search space. The long-range head consumes bird's-eye-view splatted frustums with class priors, and the short-range head consumes raw frustum points via a LiDAR-based detector, with PointPillars selected as the best performer.

What would settle it

Measure per-object depth error and 3D detection recall on the OSDaR23 test set separately for the 200-250 meter range: if the full pipeline's 3D recall at that range drops to near zero when predicted depth maps are used instead of ground-truth frustums, the 250-meter detection claim would be disproven.

Watch

Extended reading notes

Core claim

The central claim is that LiDAR guidance during training alone can enable a monocular-only pipeline to produce frustum-based 3D detections at long range. MFF distills relative depth from a diffusion-based teacher (Marigold) into a real-time DenseDepth variant, then fine-tunes a refinement network to absolute depth using inpainted LiDAR ground truth. A modified YOLOv9 with an added distance head outputs 2.5D detections, and a weighted sum of frustum centroid and estimated distance routes each frustum to a short-range LiDAR-based head or a long-range BEV head. On the OSDaR23 test set the full pipeline detects objects up to 250 meters, and the long-range head reaches 60.61 3D mAP@0.1 on KITTI cars versus 46.90 for Faraway-Frustum. The authors state that depth-map and 2.5D errors propagate into frustum quality and currently limit 3D prediction accuracy.

Load-bearing premise

The method assumes that depth maps predicted from a single image, especially beyond 200 meters, are accurate enough to build frustums that preserve the 3D structure of objects; the paper itself acknowledges that errors from 2.5D predictions and depth maps degrade frustum quality and hinder 3D predictions.

Editorial extensions

If this is right

  • If correct, camera-only systems could provide early 3D warnings at 250 meters without requiring per-vehicle LiDAR at deployment, lowering cost for railway automation.
  • The long-range head architecture transfers to automotive settings, as shown by its improvement over Faraway-Frustum's head on KITTI.
  • Distilling depth from a diffusion teacher into a real-time network offers a path to accurate absolute depth estimation without ground-truth depth labels.
  • The modular pipeline implies that improving the 2.5D distance head or the depth estimation module would directly improve downstream 3D detection quality.
  • Demonstrating long-range detection on OSDaR23 provides a benchmark for future railway perception research beyond the automotive domain.

Reading between the lines

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

  • The 250-meter ceiling may be a property of depth quality more than of the frustum architecture; reducing depth error at range could push the same pipeline toward the kilometer-scale perception trains ultimately need.
  • The scarcity of long-range training samples in OSDaR23 suggests that performance beyond 200 meters could improve substantially with more balanced annotations or synthetic augmentation.
  • The weighted-sum routing between short- and long-range heads could be replaced by a learned, end-to-end mechanism, since the paper itself notes that module errors propagate through frustumization.
  • A testable extension would apply the same LiDAR-guided depth distillation to other camera-based 3D detectors to separate the contribution of depth quality from that of the frustum design.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper presents a modular monocular 3D object detection pipeline for railway scenes, termed MFF (Monocular Faraway-Frustum). The pipeline combines a YOLOv9-based 2.5D object detector, a DenseDepth/Marigold-based monocular depth estimation module, frustum generation with a weighted centroid/distance decision rule, and separate short- and long-range 3D detection heads (evaluated with PointRCNN, PointPillars, and Part-A2 for short range; an FF-Net-style head for long range). LiDAR point clouds are used only during training to supervise depth refinement and to generate frustums for head training. The authors report module-level validation results on OSDaR23 and a KITTI comparison for the long-range head, and the abstract claims that the method detects objects up to 250 meters. The full-pipeline evaluation on the OSDaR23 test set, however, reports only 2.5D detection metrics (Table VI) and depth MAE, with 3D detections shown only qualitatively in Figure 6.

Significance. If the headline claim were quantitatively supported, the work would be a valuable contribution to long-range perception for autonomous rail, a domain with comparatively few public datasets and methods. The paper has genuine strengths: it uses a relevant real-world dataset (OSDaR23), provides component-level ablation-style comparisons, includes a KITTI benchmark for the long-range head, and openly discusses limitations such as depth-map noise and label imbalance. The modular design is clearly described and reproducible in principle. However, the central claim that the full monocular pipeline detects objects up to 250 meters is never measured end-to-end in 3D. The only quantitative full-pipeline results are 2.5D mAP/MAE and depth MAE, and the authors acknowledge both substantial per-class distance errors (buffer stop MAE 119.75 m) and significant overlap between training and test sequences. These issues directly undermine the abstract's central claim and the practical relevance of the approach.

major comments (4)
  1. [Section IV-B, Table VI and Figure 6] The abstract claims that evaluations on OSDaR23 demonstrate detecting objects up to 250 meters, but the full-pipeline evaluation contains no end-to-end 3D detection metric. Table VI reports only 2.5D mAP@0.5 and MAE, and Figure 6 is qualitative. Consequently, the central claim is not quantified for the assembled monocular pipeline at any range, let alone at 200-250 m.
  2. [Section III-D / Section IV-A.3, Table III] The long-range head is evaluated using ground-truth 2D labels and OSDaR23-derived depth maps (LiDAR/inpainted), not the monocular pipeline's predicted 2.5D boxes and estimated depth. Table III therefore measures the head in isolation under idealized inputs. The KITTI comparison in Table IV uses a different domain and class (Car) and does not validate the railway long-range setting. These numbers cannot support the claim that the assembled pipeline detects objects at 250 m.
  3. [Section IV-B, paragraph on test-set overlap] The authors state that performance on the test set closely matches validation 'given the significant overlap between training and test sequences.' This is a data-leakage concern: if test frames come from the same sequences as training frames, the reported test-set results (Table VI, depth MAE 12.85 m) are optimistic and cannot be interpreted as generalization performance. The paper does not quantify the overlap or provide results on disjoint sequences.
  4. [Section IV-B, discussion of error propagation] The authors acknowledge that 'the error carried from the 2.5D predictions and depth maps can end up affecting the quality of the frustums, which at the same time ends up hindering 3D object predictions.' Given the empirically reported errors (e.g., buffer stop MAE 119.75 m at test time, full MAE 15.76 m, depth MAE 12.85 m), the effect on long-range frustum geometry is likely severe. This admission, combined with the absence of an end-to-end 3D metric, means the 'up to 250 meters' claim is not merely unproven but is contradicted by the reported error magnitudes.
minor comments (4)
  1. [Abstract / Introduction] The phrase 'an deep-learning-based approach' should be 'a deep-learning-based approach'.
  2. [Section IV-A.3, Table III] The table caption and text would benefit from explicitly stating that the long-range head evaluation uses ground-truth 2D labels and OSDaR23 depth maps, so the reader does not mistake these for end-to-end results.
  3. [Section IV-A.4, Table V] The assessment states that PointPillars was chosen for the full pipeline, but no end-to-end comparison of the three short-range heads in the assembled pipeline is reported; a sentence explaining why the integration-level comparison was omitted would clarify the evaluation strategy.
  4. [Throughout] Several minor typographical errors and inconsistencies appear (e.g., 'utlized', 'Associationation', 'V oixel', 'GTs' vs 'GT'); a careful proofreading pass is recommended.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: each module is trained and evaluated on its own inputs; the 250 m claim is under-supported but not definitionally forced.

full rationale

The paper's derivation chain is empirical rather than analytic: the 2.5D detector, depth network, and 3D heads are separately trained on OSDaR23 labels, LiDAR-derived depth maps, and KITTI data, and the reported component metrics are measured on validation/test splits against those same external labels. No predicted quantity is defined in terms of the quantity it is later claimed to predict, and no parameter is fitted to a subset and then reported as a prediction of a closely related forced value. The long-range head is trained with ground-truth 2D labels and OSDaR23 depth maps, and its KITTI comparison against Faraway-Frustum is an external benchmark, not a self-citation. The only overlapping-author citation, [19], appears as a general example of hybrid methods in related work and is not load-bearing. The paper's genuine weakness is evaluative: the full monocular pipeline is never scored with an end-to-end 3D detection metric on OSDaR23, and Section IV-B acknowledges that 'the error carried from the 2.5D predictions and depth maps can end up affecting the quality of the frustums' and that there is 'significant overlap between training and test sequences.' These are evidence that the 250-meter claim is under-supported by the presented experiments, but they are not circularity: the pipeline is not equivalent to its inputs by construction, and no result is forced by a self-citation chain or by definition. The appropriate verdict is no significant circularity.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The central claim rests on a set of design choices and external tools. The most significant pull-ins are the LiDAR-derived depth ground truth, the Marigold teacher, and the OSDaR23 labels. The paper contributes integration and evaluation, not a derivation.

free parameters (3)
  • Long/short range split threshold = 100 m
    Section III defines distant objects as those beyond 100 meters; this threshold routes objects to different detection heads and is chosen by hand rather than learned or swept.
  • Frustum center weighted-sum coefficient = not reported
    Section III-C says frustum centroid and 2.5D distance are combined with a weighted sum, but the weight is never given; changing it changes which head processes each object.
  • Distance head output range = 0 to 250 m
    The distance head produces normalized values representing distances between 0 and 250 meters, which caps the maximum detectable distance and is a design choice.
assumptions (5)
  • domain assumption LiDAR-derived inpainted depth maps are accurate enough to serve as ground truth for absolute depth and pseudo-cloud generation.
    Section III-A trains the refinement network on inpainted depth maps produced from OSDaR23 point clouds; errors in inpainting are not quantified and propagate to frustums.
  • domain assumption Marigold's relative depth estimates are reliable teacher signals for the DenseDepth student on railway scenes.
    Section III-A uses Marigold as teacher with no validation that its relative depth is unbiased on OSDaR23 or RailSem19.
  • standard math Back-projection of depth maps to pseudo-point-clouds, and subsequent BEV splatting, preserves the information needed for 3D box regression.
    The pipeline relies on camera intrinsics and extrinsics and known projection formulas, but errors in estimated depth are not modeled.
  • domain assumption Knowledge transfers from KITTI to the railway domain for the 2.5D detection module.
    Section III-B pretrains on KITTI plus OSDaR23 and fine-tunes on OSDaR23; the paper does not analyze domain shift.
  • domain assumption ASAM OpenLABEL 3D cuboids and LiDAR points are reliable ground truth.
    Evaluation uses OSDaR23 labels without discussing annotation noise or object truncation at long range.

how reviews work

0 comments
Cite this review

Pith. "Pith review of LiDAR-Guided Monocular 3D Object Detection for Long-Range Railway Monitoring." pith.science (2026). https://pith.science/paper/RSLPB6SN

@misc{pith2026250418203,
  author       = {Pith},
  title        = {Pith review of: LiDAR-Guided Monocular 3D Object Detection for Long-Range Railway Monitoring},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RSLPB6SN}},
  note         = {Machine review of arXiv:2504.18203}
}
read the original abstract

Railway systems, particularly in Germany, require high levels of automation to address legacy infrastructure challenges and increase train traffic safely. A key component of automation is robust long-range perception, essential for early hazard detection, such as obstacles at level crossings or pedestrians on tracks. Unlike automotive systems with braking distances of ~70 meters, trains require perception ranges exceeding 1 km. This paper presents an deep-learning-based approach for long-range 3D object detection tailored for autonomous trains. The method relies solely on monocular images, inspired by the Faraway-Frustum approach, and incorporates LiDAR data during training to improve depth estimation. The proposed pipeline consists of four key modules: (1) a modified YOLOv9 for 2.5D object detection, (2) a depth estimation network, and (3-4) dedicated short- and long-range 3D detection heads. Evaluations on the OSDaR23 dataset demonstrate the effectiveness of the approach in detecting objects up to 250 meters. Results highlight its potential for railway automation and outline areas for future improvement.

Figures

Figures reproduced from arXiv: 2504.18203 by the authors.

Figure 1
Figure 1. MFF Pipeline. Top: Pipeline during training, when LiDAR point clouds are used to guide depth estimation. Bottom: [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 3
Figure 3. Depth estimation module prediction example (relative [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 2
Figure 2. Number of point cloud points within the 3D bounding [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figures from the paper (2 more)
Figure 5
Figure 5. Figure 5: Heatmap exemplifying the error distribution in the [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Examples of detections made by MFF, green boxes [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

51 extracted references · 37 canonical work pages

  1. [1]

    Trains: The backbone of mobility,

    C. Borghini, “Trains: The backbone of mobility,” Pais Positivo, 2021, originally published in Portuguese in the April edition of the Pais Positivo magazine, issue no 144

    work page 2021

  2. [2]

    What is digitale schiene deutschland?

    “What is digitale schiene deutschland?” https://digitale-schiene-deutschland.de/en/ What-is-Digitale-Schiene-Deutschland

    work page

  3. [3]

    About shift2rail,

    “About shift2rail,” https://rail-research.europa.eu/about-shift2rail/

    work page

  4. [4]

    safe.train: Rethinking mobility

    safe.trAIn, “safe.train: Rethinking mobility.” [Online]. Available: https://safetrain-projekt.de/en/

    work page

  5. [5]

    [Online]

    (2023) Railway safety statistics in the EU. [Online]. Available: https://ec.europa.eu/eurostat/statistics-explained/index.php? title=Railway safety statistics in the EU

    work page 2023

  6. [6]

    A Review of Vision-Based On-Board Obstacle Detection and Distance Estimation in Railways,

    D. Risti ´c-Durrant, M. Franke, and K. Michels, “A Review of Vision-Based On-Board Obstacle Detection and Distance Estimation in Railways,” Sensors, vol. 21, no. 10, p. 3452, 2021. [Online]. Available: https://www.mdpi.com/1424-8220/21/10/3452

    work page 2021

  7. [7]

    High-Precision Low-Cost Gimballing Platform for Long-Range Railway Obstacle Detection,

    E. H. Assaf, C. von Einem, C. Cadena, R. Siegwart, and F. Tschopp, “High-Precision Low-Cost Gimballing Platform for Long-Range Railway Obstacle Detection,” Sensors, vol. 22, no. 2, p. 474, 2022. [Online]. Available: https://www.mdpi.com/1424-8220/22/2/474

    work page 2022

  8. [8]

    Improving distant 3d object detection using 2d box supervision,

    Z. Yang, Z. Yu, C. Choy, R. Wang, A. Anandkumar, and J. M. Alvarez, “Improving distant 3d object detection using 2d box supervision,”

    work page

Show all 51 references
  1. [9]

    Lidar on its way out? camera’s market size from 76% to 79% by 2033,

    Y . Wang, “Lidar on its way out? camera’s market size from 76% to 79% by 2033,” IDTechEx research , 2023. [Online]. Available: https://www.idtechex.com/en/research-article/ lidar-on-its-way-out-cameras-market-size-from-76-to-79-by-2033/ 28865

    2023

  2. [10]

    Virtual Sparse Convolution for Multimodal 3D Object Detection,

    H. Wu, C. Wen, S. Shi, X. Li, and C. Wang, “Virtual Sparse Convolution for Multimodal 3D Object Detection,” in CVPR 2023 , 2023, pp. 21 653–21 662

    2023

  3. [11]

    LoGoNet: Towards Accurate 3D Object Detection With Local-to-Global Cross-Modal Fusion,

    X. Li, T. Ma, Y . Hou, B. Shi, Y . Yang, Y . Liu, X. Wu, Q. Chen, Y . Li, Y . Qiao, and L. He, “LoGoNet: Towards Accurate 3D Object Detection With Local-to-Global Cross-Modal Fusion,” in CVPR 2023, 2023, pp. 17 524–17 534

    2023

  4. [12]

    V oxel Field Fusion for 3D Object Detection,

    Y . Li, X. Qi, Y . Chen, L. Wang, Z. Li, J. Sun, and J. Jia, “V oxel Field Fusion for 3D Object Detection,” inCVPR 2022, 2022, pp. 1120–1129

    2022

  5. [13]

    Homogeneous Multi-modal Feature Fusion and Interaction for 3D Object Detection,

    X. Li, B. Shi, Y . Hou, X. Wu, T. Ma, Y . Li, and L. He, “Homogeneous Multi-modal Feature Fusion and Interaction for 3D Object Detection,” in ECCV 2022, S. Avidan, G. Brostow, M. Ciss ´e, G. M. Farinella, and T. Hassner, Eds. Cham: Springer Nature Switzerland, 2022, pp. 691– 707

    2022

  6. [14]

    Chapter 13 - 3D object detection and tracking,

    I. Oleksiienko and A. Iosifidis, “Chapter 13 - 3D object detection and tracking,” in Deep Learning for Robot Perception and Cognition , A. Iosifidis and A. Tefas, Eds. Academic Press, 2022, pp. 313–340

    2022

  7. [15]

    MonoDETR: Depth-guided Transformer for Monocular 3D Object Detection,

    R. Zhang, H. Qiu, T. Wang, Z. Guo, Z. Cui, Y . Qiao, H. Li, and P. Gao, “MonoDETR: Depth-guided Transformer for Monocular 3D Object Detection,” in ICCV 2023, 2023, pp. 9155–9166

    2023

  8. [16]

    Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous Driving,

    Y . You, Y . Wang, W.-L. Chao, D. Garg, G. Pleiss, B. Hariharan, M. Campbell, and K. Q. Weinberger, “Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous Driving,” Feb. 2020

    2020

  9. [17]

    BEVDepth: Acquisition of Reliable Depth for Multi-View 3D Object Detection,

    Y . Li, Z. Ge, G. Yu, J. Yang, Z. Wang, Y . Shi, J. Sun, and Z. Li, “BEVDepth: Acquisition of Reliable Depth for Multi-View 3D Object Detection,” Proceedings of the AAAI Conference on Artificial Intelli- gence, vol. 37, no. 2, pp. 1477–1485, June 2023

    2023

  10. [18]

    Towards unified 3d object detection via algorithm and data unification,

    Z. Li, X. Xu, S. Lim, and H. Zhao, “Towards unified 3d object detection via algorithm and data unification,” 2024. [Online]. Available: https://arxiv.org/abs/2402.18573

    2024 arXiv

  11. [19]

    Collision avoidance route planning for autonomous medical devices using multiple depth cameras,

    M. Mahmeen, R. D. Dominguez Sanchez, M. Friebe, M. Pech, and S. Haider, “Collision avoidance route planning for autonomous medical devices using multiple depth cameras,” IEEE Access, vol. 10, pp. 29 903–29 915, 2022

    2022

  12. [20]

    Frustum PointNets for 3D Object Detection From RGB-D Data,

    C. R. Qi, W. Liu, C. Wu, H. Su, and L. J. Guibas, “Frustum PointNets for 3D Object Detection From RGB-D Data,” in CVPR 2018 , 2018, pp. 918–927

    2018

  13. [21]

    Faraway-Frustum: Dealing with Lidar Sparsity for 3D Object Detec- tion using Fusion,

    H. Zhang, D. Yang, E. Yurtsever, K. A. Redmill, and ¨U. ¨Ozg¨uner, “Faraway-Frustum: Dealing with Lidar Sparsity for 3D Object Detec- tion using Fusion,” Mar. 2021

    2021

  14. [22]

    Pv-rcnn: Point-voxel feature set abstraction for 3d object detection,

    S. Shi, C. Guo, L. Jiang, Z. Wang, J. Shi, X. Wang, and H. Li, “Pv-rcnn: Point-voxel feature set abstraction for 3d object detection,”

  15. [23]

    Mobilenets: Efficient convolutional neural networks for mobile vision applications,

    A. G. Howard, M. Zhu, B. Chen, D. Kalenichenko, W. Wang, T. Weyand, M. Andreetto, and H. Adam, “Mobilenets: Efficient convolutional neural networks for mobile vision applications,” 2017. [Online]. Available: https://arxiv.org/abs/1704.04861

    2017 arXiv

  16. [24]

    Are we ready for autonomous driving? the kitti vision benchmark suite,

    A. Geiger, P. Lenz, and R. Urtasun, “Are we ready for autonomous driving? the kitti vision benchmark suite,” in Conference on Computer Vision and Pattern Recognition (CVPR) , 2012

    2012

  17. [25]

    You only look once: Unified, real-time object detection,

    J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, “You only look once: Unified, real-time object detection,” in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , 2016, pp. 779– 788

    2016

  18. [26]

    Yolov9: Learning what you want to learn using programmable gradient information,

    C.-Y . Wang, I.-H. Yeh, and H.-Y . M. Liao, “Yolov9: Learning what you want to learn using programmable gradient information,” 2024. [Online]. Available: https://arxiv.org/abs/2402.13616

    2024 arXiv

  19. [27]

    Yolov1 to yolov10: The fastest and most accurate real-time object detection systems,

    C.-Y . Wang and H.-Y . M. Liao, “Yolov1 to yolov10: The fastest and most accurate real-time object detection systems,” 2024. [Online]. Available: https://arxiv.org/abs/2408.09332

    2024 arXiv

  20. [28]

    Machine learning techniques for autonomous multi- sensor long-range environmental perception system,

    M. A. Haseeb, “Machine learning techniques for autonomous multi- sensor long-range environmental perception system,” Ph.D. disserta- tion, Universit¨at Bremen, 2021

    2021

  21. [29]

    Dist-yolo: Fast object detection with distance estimation,

    M. Vajgl, P. Hurtik, and T. Nejezchleba, “Dist-yolo: Fast object detection with distance estimation,” Applied Sciences, vol. 12, no. 3,

  22. [30]

    Masoumian, D

    A. Masoumian, D. G. Marei, S. Abdulwahab, J. Cristiano, D. Puig, and H. A. Rashwan, Absolute Distance Prediction Based on Deep Learning Object Detection and Monocular Depth Estimation Models . IOS Press, Oct. 2021. [Online]. Available: http://dx.doi.org/10.3233/FAIA210151

    2021 doi

  23. [31]

    Towards robust monocular depth estimation: Mixing datasets for zero-shot cross-dataset transfer,

    R. Ranftl, K. Lasinger, D. Hafner, K. Schindler, and V . Koltun, “Towards robust monocular depth estimation: Mixing datasets for zero-shot cross-dataset transfer,” 2020. [Online]. Available: https://arxiv.org/abs/1907.01341

    2020 arXiv

  24. [32]

    High quality monocular depth estimation via transfer learning,

    I. Alhashim and P. Wonka, “High quality monocular depth estimation via transfer learning,” arXiv e-prints , vol. abs/1812.11941, 2018. [Online]. Available: https://arxiv.org/abs/1812.11941

    2018 arXiv

  25. [33]

    Densely connected convolutional networks,

    G. Huang, Z. Liu, L. van der Maaten, and K. Q. Weinberger, “Densely connected convolutional networks,” 2018. [Online]. Available: https://arxiv.org/abs/1608.06993

    2018 arXiv

  26. [34]

    Colorization using optimiza- tion,

    A. Levin, D. Lischinski, and Y . Weiss, “Colorization using optimiza- tion,” in ACM SIGGRAPH 2004 Papers , 2004, pp. 689–694

    2004

  27. [35]

    Repurposing diffusion-based image generators for monocular depth estimation,

    B. Ke, A. Obukhov, S. Huang, N. Metzger, R. C. Daudt, and K. Schindler, “Repurposing diffusion-based image generators for monocular depth estimation,” in Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition (CVPR) , 2024

    2024

  28. [36]

    nuscenes: A multimodal dataset for autonomous driving,

    H. Caesar, V . Bankiti, A. H. Lang, S. V ora, V . E. Liong, Q. Xu, A. Krishnan, Y . Pan, G. Baldan, and O. Beijbom, “nuscenes: A multimodal dataset for autonomous driving,” in CVPR, 2020

    2020

  29. [37]

    Waymo open dataset: Panoramic video panoptic segmentation,

    J. Mei, A. Z. Zhu, X. Yan, H. Yan, S. Qiao, Y . Zhu, L.-C. Chen, H. Kretzschmar, and D. Anguelov, “Waymo open dataset: Panoramic video panoptic segmentation,” 2022. [Online]. Available: https://arxiv.org/abs/2206.07704

    2022 arXiv

  30. [38]

    RailSem19: A Dataset for Semantic Rail Scene Understanding,

    O. Zendel, M. Murschitz, M. Zeilinger, D. Steininger, S. Abbasi, and C. Beleznai, “RailSem19: A Dataset for Semantic Rail Scene Understanding,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops , 2019

    2019

  31. [39]

    Railnet: A segmentation network for railroad detection,

    Y . Wang, L. Wang, Y . H. Hu, and J. Qiu, “Railnet: A segmentation network for railroad detection,” IEEE Access , vol. 7, pp. 143 772– 143 779, 2019

    2019

  32. [40]

    Frsign: A large-scale traffic light dataset for autonomous trains,

    J. Harb, N. R ´eb´ena, R. Chosidow, G. Roblin, R. Potarusov, and H. Hajri, “Frsign: A large-scale traffic light dataset for autonomous trains,” 2020. [Online]. Available: https://arxiv.org/abs/2002.05665

    2020 arXiv

  33. [41]

    A lightweight framework for obstacle detection in the railway image based on fast region proposal and improved yolo-tiny network,

    L. Guan, L. Jia, Z. Xie, and C. Yin, “A lightweight framework for obstacle detection in the railway image based on fast region proposal and improved yolo-tiny network,” IEEE Transactions on Instrumentation and Measurement , vol. 71, pp. 1–16, 2022

    2022

  34. [42]

    Railset: A unique dataset for railway anomaly detection,

    A. Zouaoui, A. Mahtani, M. A. Hadded, S. Ambellouis, J. Boonaert, and H. Wannous, “Railset: A unique dataset for railway anomaly detection,” in 2022 IEEE 5th International Conference on Image Processing Applications and Systems (IPAS), vol. Five, 2022, pp. 1–6

    2022

  35. [43]

    A lightweight lidar-camera sensing method of obstacles detection and classification for autonomous rail rapid transit,

    S. Liu, C. Li, T. Yuwen, Z. Wan, and Y . Luo, “A lightweight lidar-camera sensing method of obstacles detection and classification for autonomous rail rapid transit,” IEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 12, pp. 23 043–23 058, 2022

    2022

  36. [44]

    Road and Railway Smart Mobility: A High- Definition Ground Truth Hybrid Dataset,

    R. Khemmar, A. Mauri, C. Dulompont, J. Gajula, V . Vauchey, M. Had- dad, and R. Boutteau, “Road and Railway Smart Mobility: A High- Definition Ground Truth Hybrid Dataset,” Sensors, vol. 22, no. 10, p. 3922, Jan. 2022

    2022

  37. [45]

    3D Object Detection on Synthetic Point Clouds for Railway Applications,

    M. Neri and F. Battisti, “3D Object Detection on Synthetic Point Clouds for Railway Applications,” in 2022 10th European Workshop on Visual Information Processing (EUVIP) , 2022, pp. 1–6

    2022

  38. [46]

    Open sensor data for rail 2023,

    R. Tilly, P. Neumaier, K. Schwalbe, P. Klasek, R. Tagiew, P. Denzler, T. Klockau, M. Boekhoff, and M. K ¨oppel, “Open sensor data for rail 2023,” 2023. [Online]. Available: https: //data.fid-move.de/dataset/3d7e7406-639f-49f6-bbca-caac511b4032

    2023

  39. [47]

    Asam openlabel,

    Advanced Data Controls Corp., Annotell AB, Ansys Inc., Deepen AI, Deutsches Zentrum f ¨ur Luft- und Raumfahrt e. V ., Five, iASYS Technology Solutions Pvt. Ltd., LiangDao GmbH, Peak Solution GmbH, SAIC Motor Corporation Ltd., Tata Consultancy Services Pvt. Ltd, understandAI Gm...

    2021

  40. [48]

    Multi-view 3d object detection network for autonomous driving,

    X. Chen, H. Ma, J. Wan, B. Li, and T. Xia, “Multi-view 3d object detection network for autonomous driving,” in 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017, pp. 6526– 6534

    2017

  41. [2021]

    Available: https://arxiv.org/abs/1912.13192

    [Online]. Available: https://arxiv.org/abs/1912.13192

    1912 arXiv

  42. [2022]

    Available: https://www.mdpi.com/2076-3417/12/3/ 1354

    [Online]. Available: https://www.mdpi.com/2076-3417/12/3/ 1354

  43. [2024]

    Available: https://arxiv.org/abs/2403.09230

    [Online]. Available: https://arxiv.org/abs/2403.09230

    arXiv

Pith tools

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