Pith. sign in

REVIEW 4 major objections 4 minor 58 references

The paper claims that a 3D object detector trained only on synthetic LiDAR from a digital twin beats the same model trained on real data on the real test set.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

A digital twin of a real intersection can generate LiDAR training data that matches the target location, and a detector trained on it reported 4.8% higher car AP than a model trained on real data, though with more synthetic frames.

T0 review reviewed 2026-08-05 challenge →

load-bearing objection Solid pipeline and public dataset, but the headline sim2real gain is confounded by a training-set size difference the authors themselves acknowledge. the 4 major comments →

arxiv 2509.02904 v1 pith:ZUMB6EAN submitted 2025-09-03 cs.CV

High-Fidelity Digital Twins for Bridging the Sim2Real Gap in LiDAR-Based ITS Perception

Muhammad Shahbaz , Shaurya Agarwal This is my paper
classification cs.CV
keywords digital twinLiDARsim-to-real3D object detectionCARLALUMPIdomain alignmentintelligent transportation systems
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to establish that building a high-fidelity digital twin of a real roadside scene, matching background geometry, road topology, traffic mix, and sensor specifications, makes synthetic LiDAR data effectively in-domain for training a real-world 3D detector. As evidence, a SEED detector trained only on the synthetic UT-LUMPI dataset achieves 44.74% car AP@IoU=0.5 on real LUMPI test frames, compared with 42.70% for the same model trained on real data. The paper quantifies the reduced domain gap with Chamfer Distance, Maximum Mean Discrepancy, Earth Mover's Distance, and Frechet Distance at both raw-input and latent-feature levels, showing the synthetic data aligns most closely with its target real dataset. If correct, this points to a scalable route to labeled LiDAR training data for intersections without expensive manual annotation.

Core claim

The core claim is that a location-specific digital twin can replace real labeled data for training LiDAR perception. The authors construct UT-LUMPI, a CARLA simulation of the LUMPI intersection built from a satellite 3D mesh for static geometry, OpenStreetMap-derived lane-level roads, stochastic traffic matched to the real class distribution, and virtual sensors replicating the real Hesai and Velodyne specifications and poses. Trained on 10,000 synthetic frames, an off-the-shelf SEED detector reaches car AP@IoU=0.5 of 44.74% on real LUMPI test frames, versus 42.70% for the identical model trained on real training data, a 4.8% relative improvement. Training-loss trajectories align closely, an

What carries the argument

The load-bearing mechanism is the digital twin construction pipeline: a satellite 3D mesh is cut and rescaled to the region of interest, fused with OpenStreetMap-derived road topology, populated with CARLA traffic matched to the target class distribution, and scanned by CARLA LiDAR sensors configured to the real sensors' channel count, field of view, range, point rate, and pose. The work this does is to make the simulated distribution approximate the real distribution directly in point-cloud space; the distribution-alignment metrics and latent-feature overlap confirm that the two domains occupy the same feature regions for the target location.

Load-bearing premise

CARLA's virtual LiDAR sensor is a faithful enough proxy for the real Hesai and Velodyne sensors that the point-cloud distributions match, even though CARLA does not model reflectance, multiple returns, ray dropout, or motion distortion.

What would settle it

Train the same detector on synthetic frames from the same digital twin but with the virtual sensor replaced by a generic CARLA sensor, or with the background geometry replaced by a generic CARLA town while keeping sensor specifications and traffic identical. If real LUMPI AP remains near 44.74%, the claimed fidelity mechanism is not doing the work; if AP falls toward the roughly 68% performance decline reported for naive CARLA-to-KITTI transfer, then the specific digital-twin alignment is what carries the result.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Synthetic-only training can match or exceed real-data training for a roadside 3D detection benchmark, shifting the labeling burden from manual annotation to environment construction.
  • The method produces in-domain data without post-hoc domain adaptation, unlike self-training or GAN-based correction pipelines.
  • Distributional alignment is target-specific: UT-LUMPI is far closer to LUMPI than to V2X-Real or A9, implying digital twins are built for a specific deployment site rather than being universally transferable.
  • The released synthetic datasets provide a benchmark for further sim-to-real LiDAR research and for augmenting existing real datasets.
  • Training-loss convergence similarity suggests that task-relevant statistical properties, not just visual resemblance, are what the detector uses to generalize.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The 4.8% gain may partly reflect that synthetic labels are noise-free and the synthetic set is about 23% larger than the real training set; an ablation holding sample count and label noise constant would isolate the fidelity contribution.
  • The sensor-modeling assumption is the fragile link: if CARLA's omission of reflectance, multiple returns, ray dropout, or motion distortion matters more for other sensor placements or object classes such as pedestrians and cyclists, the margin could reverse.
  • Strong alignment with LUMPI and weak alignment with other datasets suggests the pipeline's value is location-specific; applying it to a new city requires rebuilding the twin, so scalability depends on automating mesh cleanup and road fitting.
  • A testable extension would degrade the digital twin in controlled ways, such as replacing the satellite geometry with a generic CARLA town while keeping sensor specs and traffic identical, to measure how much of the detection gain comes from geometry fidelity as opposed to sensor and traffic modeling.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. This paper proposes a high-fidelity digital twin (HiFi DT) framework for generating synthetic LiDAR data for roadside ITS perception. The authors construct a CARLA replica of the LUMPI intersection using publicly available mesh imagery, OpenStreetMap road topology, and sensor specifications/pose, then generate 10,000 labeled synthetic frames (UT-LUMPI). A SEED 3D object detector trained only on these synthetic frames is evaluated on real LUMPI test frames and compared with an identical model trained on real LUMPI data. The synthetic-trained model achieves car AP@IoU=0.5 of 44.74% versus 42.70% for the real-trained model, a 4.8% relative improvement. The paper also reports Chamfer Distance, MMD, EMD, and Fréchet Distance between synthetic and real distributions at raw-input and latent-feature levels, together with t-SNE/UMAP visualizations, arguing that the twin produces in-domain data. The authors explicitly note that the synthetic dataset is larger than the real one and attribute part of the observed gain to this size difference.

Significance. The paper addresses an important and timely problem: reducing the sim2real gap for LiDAR-based ITS perception. If the headline result were obtained under controlled conditions, it would be a strong demonstration that high-fidelity digital twins can generate training data competitive with or superior to real data, with substantial cost and scalability benefits. The public release of UT-LUMPI and companion datasets is a concrete asset. The systematic pipeline and the multi-metric distribution analysis are useful contributions. However, the current experimental design does not isolate the effect of twin fidelity from dataset size, and the distributional metrics are computed on the same target site used to construct the twin. These issues must be addressed before the central claim can be accepted.

major comments (4)
  1. [Section IV, Fig. 5] The headline comparison (44.74% vs 42.70% AP@IoU=0.5) is not controlled for training-set size. UT-LUMPI has 10,000 frames and the LUMPI subset has 8,120; with an 80/20 split, the synthetic training set is 8,000 frames versus 6,496 real frames, a ~23% advantage for the synthetic condition. The paper itself concedes this in the Fig. 4 caption ('slightly more training samples, leading to better convergence behavior') and in Section IV ('This slight performance gain can be attributed to the marginally more training samples available in the synthetic dataset'). This directly undermines the causal attribution of the gain to HiFi DT fidelity. Please rerun with equal training-set sizes (e.g., subsample synthetic to match real count) and report multiple seeds with mean and standard deviation.
  2. [Section IV, Fig. 8 and Section III-B] The distributional alignment metrics (CD=0.32, MMD=1.05e-5, EMD=0.988, FD=0.210 in Fig. 8) compare UT-LUMPI to LUMPI, which is the exact site whose geometry, road topology, and sensor pose were used to construct the twin. This is partly a self-consistency check, not independent evidence that the HiFi DT approach reduces domain shift. The comparison to V2X-Real and A9 only shows that different geographic sites differ. Add a control simulation in a generic CARLA town, or a twin with one component removed (e.g., generic road topology or default sensor pose), and show that the same metrics and downstream AP degrade accordingly.
  3. [Section II and Table I] The related-work section correctly states that simulators such as CARLA and LGSVL do not model material reflectance, ray dropout on glass, multiple returns, or motion distortion. Yet the virtual sensors in Table I are configured with only nominal specifications (channels, FOV, points/sec, range), and no experiment quantifies how these missing physical effects influence point-cloud alignment or detector performance. The claim that the generated point clouds are 'in-domain' would be substantially strengthened by a sensor-model sensitivity analysis, or by explicit evidence that these effects are negligible for the roadside geometry and sensor range considered here.
  4. [Section IV, Training Deep Object Detector] No repeated training runs are reported. A 2.04-point AP difference between a single synthetic-trained and a single real-trained model is within plausible run-to-run variance for a DETR-style detector with 12 epochs. Reporting only one run per condition makes it impossible to assess whether the observed difference is statistically meaningful. Please provide results over multiple random seeds (at least 3), with mean and standard deviation, and state whether the difference is significant.
minor comments (4)
  1. [Abstract and Section I] Typo: 'inclding' should be 'including'. The symbol 'Fréchet Distance' is malformed in several places (e.g., Abstract, Section III-A).
  2. [Section IV, Fig. 4 caption] Calling a ~23% training-size difference 'slightly more' is misleading. Please use exact frame counts or percentages.
  3. [Section III-B, step 5] The text says 'sensor scarification' (should be 'specifications'). Sensor pose is described as center and height, but Table I does not report tilt, which is a free parameter in the alignment; please clarify how tilt was set and whether it was tuned on real data.
  4. [Section I-A, first contribution bullet] The claim of being the 'first study' is too strong given the prior digital-twin lidar works cited in Section II (e.g., [52], [53]). Please qualify the novelty claim.

Circularity Check

0 steps flagged

No significant circularity: the headline AP comparison is empirical and the distribution metrics are descriptive; the main caveat is a confound, not a circular derivation.

full rationale

The paper's derivation chain is empirical rather than deductive. A CARLA digital twin of the LUMPI site is constructed using public map geometry, sensor specifications (Table I), and sensor poses estimated from one real LUMPI frame; 10,000 synthetic labeled frames are generated; an off-the-shelf SEED detector is trained on those frames; and the detector is evaluated on the real LUMPI test split. The synthetic labels are generated independently of the real labels, and the test split is held out, so the core AP comparison (44.74% vs 42.70%) is not circular. The distribution-alignment metrics (CD, MMD, EMD, FD, t-SNE/UMAP) are descriptive evaluations of the construction, not predictions derived from the construction; showing that a digital twin built to match LUMPI resembles LUMPI is a self-consistency check, but it is not a reduction of an output to an input. The paper's self-citations (UT-LUMPI [32], UT-V2X-Real [33], UT-TUMTraf-I [34]) are data pointers and are not load-bearing. The main methodological weakness—acknowledged by the paper itself in the Fig. 4 caption and Section IV—is that the synthetic training set was 'deliberately made slightly larger' than the real set (10,000 vs 8,120 frames), and the paper attributes the 2.04-point AP gap to this size difference; this is a confound affecting attribution of the gain to fidelity, not a circularity. Similarly, using a real LUMPI frame to estimate sensor pose is a calibration step that could leak test-set geometry if that frame is in the test split, but no specific reduction of the claimed result to its inputs can be exhibited. Under the stated rules requiring an explicit equation-level reduction or a fitted parameter renamed as a prediction, no circular step is identifiable.

Axiom & Free-Parameter Ledger

2 free parameters · 3 axioms · 0 invented entities

The pipeline rests on a few externally sourced components (CARLA, Cesium mesh, OSM) and a small set of fitted values (sensor pose, traffic mix). No new physical entities are introduced. The main uncertainty is the fidelity of the sensor simulation.

free parameters (2)
  • Sensor pose (center and tilt) = estimated from one real LUMPI frame via visual queues (concentric ground rings)
    Used to place virtual sensors in the twin; exact pose is not measured, only visually approximated, so there is no uncertainty estimate.
  • Traffic spawn distribution = not specified in the paper, described only as 'match, statistically, the distribution of target real location'
    The synthetic scene complexity and object volume stats differ from real (49.3 vs 44.2 boxes/frame, 12.2 vs 10.1 m^3) showing the traffic generation is tuned approximately.
axioms (3)
  • standard math Domain adaptation theory (Ben-David et al., ref [35]) bounds expected real loss by source loss plus domain divergence.
    Used to define the objective in Section III-A, equation (2); no new derivation is provided.
  • domain assumption CARLA's lidar sensor simulation is sufficiently realistic to produce point clouds in-domain to the target real sensors.
    The paper relies on CARLA's built-in sensor model with configured specs (Table I) but does not model reflectance, multiple returns, or motion distortion; this is a load-bearing simplification.
  • domain assumption Satellite 3D mesh (Cesium) and OpenStreetMap road topology accurately represent real-world background geometry after manual cleanup and rescaling.
    The static scene is built from these public sources; errors in geometry or alignment directly affect the synthetic point clouds.

reviewed 2026-08-05 · how reviews work

0 comments
Cite this review

Pith. "Pith review of High-Fidelity Digital Twins for Bridging the Sim2Real Gap in LiDAR-Based ITS Perception." pith.science (2026). https://pith.science/paper/ZUMB6EAN

@misc{pith2026250902904,
  author       = {Pith},
  title        = {Pith review of: High-Fidelity Digital Twins for Bridging the Sim2Real Gap in LiDAR-Based ITS Perception},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZUMB6EAN}},
  note         = {Machine review of arXiv:2509.02904}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Sim2Real domain transfer offers a cost-effective and scalable approach for developing LiDAR-based perception (e.g., object detection, tracking, segmentation) in Intelligent Transportation Systems (ITS). However, perception models trained in simulation often under perform on real-world data due to distributional shifts. To address this Sim2Real gap, this paper proposes a high-fidelity digital twin (HiFi DT) framework that incorporates real-world background geometry, lane-level road topology, and sensor-specific specifications and placement. We formalize the domain adaptation challenge underlying Sim2Real learning and present a systematic method for constructing simulation environments that yield in-domain synthetic data. An off-the-shelf 3D object detector is trained on HiFi DT-generated synthetic data and evaluated on real data. Our experiments show that the DT-trained model outperforms the equivalent model trained on real data by 4.8%. To understand this gain, we quantify distributional alignment between synthetic and real data using multiple metrics, including Chamfer Distance (CD), Maximum Mean Discrepancy (MMD), Earth Mover's Distance (EMD), and Fr'echet Distance (FD), at both raw-input and latent-feature levels. Results demonstrate that HiFi DTs substantially reduce domain shift and improve generalization across diverse evaluation scenarios. These findings underscore the significant role of digital twins in enabling reliable, simulation-based LiDAR perception for real-world ITS applications.

Figures

Figures reproduced from arXiv: 2509.02904 by Muhammad Shahbaz, Shaurya Agarwal.

Figure 1
Figure 1. Figure 1: A qualitative comparison between real scene, our high-fidelity digital twin, and generic built-in CARLA environment. In our modeling, numerous [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Systematic Method to Model High-Fidelity Digital Twin for Real-World Locations. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Frame-Level Statistical Comparison of Synthetic and Real Lidar Datasets: The alignment in point density emerges due to accurate sensor modeling. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Training loss comparison between real (LUMPI [11]) and synthetic (UT-LUMPI [32]) datasets. The plots illustrate the behavior of major loss [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Comparison of detection accuracy of SEED [30] models trained on real [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: t-SNE visualization of SEED backbone features comparing the synthetic high-fidelity digital-twin-based dataset (UT-LUMPI) to target (LUMPI) and [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: UMAP-based comparison of feature distributions between the HiFi-DT-based dataset and three real-world datasets: LUMPI (target location for HiFi-DT [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Quantitative comparison of distributional similarity between HiFi-DT-based and real data features using geometric and latent-feature metrics. The [PITH_FULL_IMAGE:figures/full_fig_p009_8.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

58 extracted references · 54 canonical work pages · 2 internal anchors

  1. [1]

    Phase diagram in multi-phase heterogeneous traffic flow model integrating the perceptual range difference under human-driven and connected vehicles environ- ment,

    C. Zhai, K. Li, R. Zhang, T. Peng, and C. Zong, “Phase diagram in multi-phase heterogeneous traffic flow model integrating the perceptual range difference under human-driven and connected vehicles environ- ment,” Chaos, Solitons & Fractals, vol. 182, p. 114 791, 2024

    work page 2024

  2. [2]

    Vision meets robotics: The kitti dataset,

    A. Geiger, P. Lenz, C. Stiller, and R. Urtasun, “Vision meets robotics: The kitti dataset,” The international journal of robotics research, vol. 32, no. 11, pp. 1231– 1237, 2013

    work page 2013

  3. [3]

    Scalability in perception for autonomous driving: Waymo open dataset,

    P. Sun, H. Kretzschmar, X. Dotiwalla, et al., “Scalability in perception for autonomous driving: Waymo open dataset,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2020, pp. 2446–2454

    work page 2020

  4. [4]

    Nuscenes: A multimodal dataset for autonomous driving,

    H. Caesar, V . Bankiti, A. H. Lang, et al. , “Nuscenes: A multimodal dataset for autonomous driving,” in Pro- ceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition , 2020, pp. 11 621–11 631

    work page 2020

  5. [5]

    Argov- erse: 3d tracking and forecasting with rich maps,

    M.-F. Chang, J. Lambert, P. Sangkloy, et al. , “Argov- erse: 3d tracking and forecasting with rich maps,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2019, pp. 8748–8757

    work page 2019

  6. [6]

    A2d2: Audi autonomous driving dataset,

    J. Geyer, Y . Kassahun, M. Mahmudi, et al. , “A2d2: Audi autonomous driving dataset,” arXiv preprint arXiv:2004.06320, 2020

    Pith/arXiv arXiv 2004

  7. [7]

    The h3d dataset for full-surround 3d multi-object detection and tracking in crowded urban scenes,

    A. Patil, S. Malla, H. Gang, and Y .-T. Chen, “The h3d dataset for full-surround 3d multi-object detection and tracking in crowded urban scenes,” in 2019 Interna- tional Conference on Robotics and Automation (ICRA) , IEEE, 2019, pp. 9552–9557

    work page 2019

  8. [8]

    One million scenes for autonomous driving: Once dataset,

    J. Mao, M. Niu, C. Jiang, et al. , “One million scenes for autonomous driving: Once dataset,” arXiv preprint arXiv:2106.11037, 2021

    Pith/arXiv arXiv 2021

  9. [9]

    Pandaset: Advanced sensor suite dataset for autonomous driving,

    P. Xiao, Z. Shao, S. Hao, et al. , “Pandaset: Advanced sensor suite dataset for autonomous driving,” in 2021 IEEE international intelligent transportation systems conference (ITSC), IEEE, 2021, pp. 3095–3101

    work page 2021

  10. [10]

    Ips300+: A challeng- ing multi-modal data sets for intersection perception system,

    H. Wang, X. Zhang, Z. Li, et al., “Ips300+: A challeng- ing multi-modal data sets for intersection perception system,” in 2022 International Conference on Robotics and Automation (ICRA) , IEEE, 2022, pp. 2539–2545

    work page 2022

  11. [11]

    Lumpi: The leibniz university multi-perspective in- tersection dataset,

    S. Busch, C. Koetsier, J. Axmann, and C. Brenner, “Lumpi: The leibniz university multi-perspective in- tersection dataset,” in 2022 IEEE Intelligent Vehicles Symposium (IV), IEEE, 2022, pp. 1127–1134

    work page 2022

  12. [12]

    Tumtraf intersection dataset: All you need for urban 3d camera-lidar roadside perception,

    W. Zimmer, C. Creß, H. T. Nguyen, and A. C. Knoll, “Tumtraf intersection dataset: All you need for urban 3d camera-lidar roadside perception,” in 2023 IEEE 26th International Conference on Intelligent Transportation Systems (ITSC), IEEE, 2023, pp. 1030–1037

    work page 2023

  13. [13]

    R-LiViT: A LiDAR-Visual-Thermal Dataset Enabling Vulnerable Road User Focused Roadside Perception

    J. Mirlach, L. Wan, A. Wiedholz, H. E. Keen, and A. Eich, “R-livit: A lidar-visual-thermal dataset en- abling vulnerable road user focused roadside percep- tion,” arXiv preprint arXiv:2503.17122 , 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  14. [14]

    A9-dataset: Multi-sensor infrastructure-based dataset for mobility research,

    C. Creß, W. Zimmer, L. Strand, et al. , “A9-dataset: Multi-sensor infrastructure-based dataset for mobility research,” in 2022 IEEE Intelligent Vehicles Symposium (IV), IEEE, 2022, pp. 965–970

    work page 2022

  15. [15]

    Rope3d: The roadside perception dataset for autonomous driving and monoc- ular 3d object detection task,

    X. Ye, M. Shu, H. Li, et al. , “Rope3d: The roadside perception dataset for autonomous driving and monoc- ular 3d object detection task,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 21 341–21 350

    work page 2022

  16. [16]

    Dair-v2x: A large-scale dataset for vehicle-infrastructure cooperative 3d object detection,

    H. Yu, Y . Luo, M. Shu, et al., “Dair-v2x: A large-scale dataset for vehicle-infrastructure cooperative 3d object detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2022, pp. 21 361–21 370

    work page 2022

  17. [17]

    V2x-real: A largs- scale dataset for vehicle-to-everything cooperative per- ception,

    H. Xiang, Z. Zheng, X. Xia, et al., “V2x-real: A largs- scale dataset for vehicle-to-everything cooperative per- ception,” in European Conference on Computer Vision , Springer, 2024, pp. 455–470

    work page 2024

  18. [18]

    Baai- vanjee roadside dataset: Towards the connected auto- mated vehicle highway technologies in challenging en- vironments of china,

    D. Yongqiang, W. Dengjiang, C. Gang, et al. , “Baai- vanjee roadside dataset: Towards the connected auto- mated vehicle highway technologies in challenging en- vironments of china,” arXiv preprint arXiv:2105.14370, 2021

    Pith/arXiv arXiv 2021

  19. [19]

    V2v4real: A real-world large-scale dataset for vehicle-to-vehicle cooperative perception,

    R. Xu, X. Xia, J. Li, et al. , “V2v4real: A real-world large-scale dataset for vehicle-to-vehicle cooperative perception,” in Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition , 2023, pp. 13 712–13 722

    work page 2023

  20. [20]

    The highd dataset: A drone dataset of naturalistic vehicle trajectories on german highways for validation of highly automated driving systems,

    R. Krajewski, J. Bock, L. Kloeker, and L. Eckstein, “The highd dataset: A drone dataset of naturalistic vehicle trajectories on german highways for validation of highly automated driving systems,” in 2018 21st international conference on intelligent transportation systems (ITSC), IEEE, 2018, pp. 2118–2125

    work page 2018

  21. [21]

    Choose your sim- ulator wisely: A review on open-source simulators for autonomous driving,

    Y . Li, W. Yuan, S. Zhang, et al. , “Choose your sim- ulator wisely: A review on open-source simulators for autonomous driving,” IEEE Transactions on Intelligent Vehicles, 2024

    work page 2024

  22. [22]

    Carla: An open urban driving simulator,

    A. Dosovitskiy, G. Ros, F. Codevilla, A. Lopez, and V . Koltun, “Carla: An open urban driving simulator,” in Conference on robot learning , PMLR, 2017, pp. 1–16

    work page 2017

  23. [23]

    Team, Deepdrive: A simulator that allows anyone with a pc to push the state-of-the-art in self-driving , 2020

    D. Team, Deepdrive: A simulator that allows anyone with a pc to push the state-of-the-art in self-driving , 2020

    work page 2020

  24. [24]

    Vista 2.0: An open, data-driven simulator for multimodal sensing and policy learning for autonomous vehicles,

    A. Amini, T.-H. Wang, I. Gilitschenski, et al. , “Vista 2.0: An open, data-driven simulator for multimodal sensing and policy learning for autonomous vehicles,” in 2022 International Conference on Robotics and Au- tomation (ICRA), IEEE, 2022, pp. 2419–2426

    work page 2022

  25. [25]

    How much real data do we actually need: Analyzing object detection performance using synthetic and real data

    F. E. Nowruzi, P. Kapoor, D. Kolhatkar, F. A. Hassanat, R. Laganiere, and J. Rebut, “How much real data do we actually need: Analyzing object detection perfor- mance using synthetic and real data,” arXiv preprint arXiv:1907.07061, 2019. JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 11

    work page internal anchor Pith review Pith/arXiv arXiv 1907

  26. [26]

    Lidar- sim: Realistic lidar simulation by leveraging the real world,

    S. Manivasagam, S. Wang, K. Wong, et al. , “Lidar- sim: Realistic lidar simulation by leveraging the real world,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2020, pp. 11 167–11 176

    work page 2020

  27. [27]

    Performance of lidar object detection deep learning architectures based on artificially generated point cloud data from carla simula- tor,

    D. Dworak, F. Ciepiela, J. Derbisz, I. Izzat, M. Ko- morkiewicz, and M. W ´ojcik, “Performance of lidar object detection deep learning architectures based on artificially generated point cloud data from carla simula- tor,” in 2019 24th International Conference on Methods and Models in Automation and Robotics (MMAR), 2019, pp. 600–605. DOI: 10.1109/MMAR.20...

    work page arXiv 2019

  28. [28]

    A survey on deep domain adaptation for lidar per- ception,

    L. T. Triess, M. Dreissig, C. B. Rist, and J. M. Z ¨ollner, “A survey on deep domain adaptation for lidar per- ception,” in 2021 IEEE intelligent vehicles symposium workshops (IV workshops) , IEEE, 2021, pp. 350–357

    work page 2021

  29. [29]

    openstreetmap.org, 2017

    OpenStreetMap contributors, Planet dump retrieved from https://planet.osm.org , https : / / www . openstreetmap.org, 2017

    work page 2017

  30. [30]

    Seed: A simple and effective 3d detr in point clouds,

    Z. Liu, J. Hou, X. Ye, T. Wang, J. Wang, and X. Bai, “Seed: A simple and effective 3d detr in point clouds,” in European Conference on Computer Vision, Springer, 2024, pp. 110–126

    work page 2024

  31. [31]

    Umap: Uniform manifold approximation and projection for dimension reduction,

    L. McInnes, J. Healy, and J. Melville, “Umap: Uniform manifold approximation and projection for dimension reduction,” arXiv preprint arXiv:1802.03426 , 2018

    Pith/arXiv arXiv 2018

  32. [32]

    Shahbaz and S

    M. Shahbaz and S. Agarwal, UT-LUMPI, version V1,

    work page

  33. [33]

    Shahbaz and S

    M. Shahbaz and S. Agarwal, UT-V2X-Real-IC, ver- sion V2, 2025. DOI: 10.7910/DVN/N6N4UR. [Online]. Available: https://doi.org/10.7910/DVN/N6N4UR

    work page doi:10.7910/dvn/n6n4ur 2025

  34. [34]

    Shahbaz and S

    M. Shahbaz and S. Agarwal, UT-TUMTraf-I, ver- sion V2, 2025. DOI: 10.7910/DVN/D21HNZ. [Online]. Available: https://doi.org/10.7910/DVN/D21HNZ

    work page doi:10.7910/dvn/d21hnz 2025

  35. [35]

    A theory of learning from different domains,

    S. Ben-David, J. Blitzer, K. Crammer, A. Kulesza, F. Pereira, and J. W. Vaughan, “A theory of learning from different domains,” Machine learning, vol. 79, pp. 151– 175, 2010

    work page 2010

  36. [36]

    Train in germany, test in the usa: Making 3d object detectors gener- alize,

    Y . Wang, X. Chen, Y . You, et al. , “Train in germany, test in the usa: Making 3d object detectors gener- alize,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2020, pp. 11 713–11 723

    work page 2020

  37. [37]

    See eye to eye: A lidar-agnostic 3d detection frame- work for unsupervised multi-target domain adaptation,

    D. Tsai, J. S. Berrio, M. Shan, S. Worrall, and E. Nebot, “See eye to eye: A lidar-agnostic 3d detection frame- work for unsupervised multi-target domain adaptation,” IEEE Robotics and Automation Letters , vol. 7, no. 3, pp. 7904–7911, 2022

    work page 2022

  38. [38]

    Performance of lidar ob- ject detection deep learning architectures based on arti- ficially generated point cloud data from carla simulator,

    D. Dworak, F. Ciepiela, J. Derbisz, I. Izzat, M. Ko- morkiewicz, and M. W ´ojcik, “Performance of lidar ob- ject detection deep learning architectures based on arti- ficially generated point cloud data from carla simulator,” in 2019 24th International Conference on Methods and Models in Automation and Robotics (MMAR) , IEEE, 2019, pp. 600–605

    work page 2019

  39. [39]

    Augmented lidar simulator for autonomous driving,

    J. Fang, D. Zhou, F. Yan, et al. , “Augmented lidar simulator for autonomous driving,” IEEE Robotics and Automation Letters, vol. 5, no. 2, pp. 1931–1938, 2020

    work page 1931

  40. [40]

    A lidar point cloud generator: From a virtual world to autonomous driving,

    X. Yue, B. Wu, S. A. Seshia, K. Keutzer, and A. L. Sangiovanni-Vincentelli, “A lidar point cloud generator: From a virtual world to autonomous driving,” in Pro- ceedings of the 2018 ACM on international conference on multimedia retrieval , 2018, pp. 458–464

    work page 2018

  41. [41]

    Using simulation data from gaming environments for training a deep learning algo- rithm on 3d point clouds,

    S. Spiegel and J. Chen, “Using simulation data from gaming environments for training a deep learning algo- rithm on 3d point clouds,” ISPRS Annals of the Pho- togrammetry, Remote Sensing and Spatial Information Sciences, vol. 8, pp. 67–74, 2021

    work page 2021

  42. [42]

    Learning to simulate realistic lidars,

    B. Guillard, S. Vemprala, J. K. Gupta, et al., “Learning to simulate realistic lidars,” in 2022 IEEE/RSJ Inter- national Conference on Intelligent Robots and Systems (IROS), IEEE, 2022, pp. 8173–8180

    work page 2022

  43. [43]

    Lgsvl simulator: A high fidelity simulator for autonomous driving,

    G. Rong, B. H. Shin, H. Tabatabaee, et al. , “Lgsvl simulator: A high fidelity simulator for autonomous driving,” in2020 IEEE 23rd International conference on intelligent transportation systems (ITSC) , IEEE, 2020, pp. 1–6

    work page 2020

  44. [44]

    Garchingsim: An autonomous driving simulator with photorealistic scenes and minimalist workflow,

    L. Zhou, Y . Song, Y . Gao, et al. , “Garchingsim: An autonomous driving simulator with photorealistic scenes and minimalist workflow,” in 2023 IEEE 26th Interna- tional Conference on Intelligent Transportation Systems (ITSC), IEEE, 2023, pp. 4227–4232

    work page 2023

  45. [45]

    Devel- opment of high-fidelity automotive lidar sensor model with standardized interfaces,

    A. Haider, M. Pigniczki, M. H. K ¨ohler, et al., “Devel- opment of high-fidelity automotive lidar sensor model with standardized interfaces,” Sensors, vol. 22, no. 19, p. 7556, 2022. DOI: 10.3390/s22197556

    work page doi:10.3390/s22197556 2022

  46. [46]

    Squeeze- seg: Convolutional neural nets with recurrent crf for real-time road-object segmentation from 3d lidar point cloud,

    B. Wu, A. Wan, X. Yue, and K. Keutzer, “Squeeze- seg: Convolutional neural nets with recurrent crf for real-time road-object segmentation from 3d lidar point cloud,” in 2018 IEEE international conference on robotics and automation (ICRA), IEEE, 2018, pp. 1887– 1893

    work page 2018

  47. [47]

    Squeezesegv2: Improved model structure and unsuper- vised domain adaptation for road-object segmentation from a lidar point cloud,

    B. Wu, X. Zhou, S. Zhao, X. Yue, and K. Keutzer, “Squeezesegv2: Improved model structure and unsuper- vised domain adaptation for road-object segmentation from a lidar point cloud,” in 2019 international confer- ence on robotics and automation (ICRA) , IEEE, 2019, pp. 4376–4382

    work page 2019

  48. [48]

    Epointda: An end- to-end simulation-to-real domain adaptation framework for lidar point cloud segmentation,

    S. Zhao, Y . Wang, B. Li, et al. , “Epointda: An end- to-end simulation-to-real domain adaptation framework for lidar point cloud segmentation,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 35, 2021, pp. 3500–3509

    work page 2021

  49. [49]

    Se- mantickitti: A dataset for semantic scene understanding of lidar sequences,

    J. Behley, M. Garbade, A. Milioto, et al. , “Se- mantickitti: A dataset for semantic scene understanding of lidar sequences,” in Proceedings of the IEEE/CVF international conference on computer vision , 2019, pp. 9297–9307

    work page 2019

  50. [50]

    CoLiGen: A unified generative framework for realistic lidar simulation in autonomous driving,

    H. Haghighi, M. Dianati, V . Donzella, and K. Debat- tista, “CoLiGen: A unified generative framework for realistic lidar simulation in autonomous driving,” IEEE Trans. Intell. Vehicles, 2024. JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 12

    work page 2024

  51. [51]

    St3d: Self-training for unsupervised domain adaptation on 3d object detection,

    J. Yang, S. Shi, Z. Wang, H. Li, and X. Qi, “St3d: Self-training for unsupervised domain adaptation on 3d object detection,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2021, pp. 10 368–10 378

    work page 2021

  52. [52]

    Digital twin-assisted graph matching multi-task object detection method in complex traffic scenarios,

    M. Li, C. Liu, X. Pan, and Z. Li, “Digital twin-assisted graph matching multi-task object detection method in complex traffic scenarios,” Scientific Reports , vol. 15, no. 1, p. 10 847, 2025

    work page 2025

  53. [53]

    Cross- dataset generalization: Bridging the gap between real and synthetic lidar data,

    M. Strunz, R. Protzmann, and I. Radusch, “Cross- dataset generalization: Bridging the gap between real and synthetic lidar data,” in International Conference on Simulation Tools and Techniques , Springer, 2024, pp. 202–217

    work page 2024

  54. [54]

    Mi- croscopic traffic simulation using sumo,

    P. A. Lopez, M. Behrisch, L. Bieker-Walz, et al., “Mi- croscopic traffic simulation using sumo,” in The 21st IEEE International Conference on Intelligent Trans- portation Systems , IEEE, 2018. [Online]. Available: https://elib.dlr.de/124092/

    work page 2018

  55. [55]

    The MathWorks, Roadrunner version: R2024b , Nat- ick, Massachusetts, United States, 2024

    I. The MathWorks, Roadrunner version: R2024b , Nat- ick, Massachusetts, United States, 2024. [Online]. Available: https : / / www . mathworks . com / products / roadrunner.html

    work page 2024

  56. [56]

    O. D. Team, Openpcdet: An open-source toolbox for 3d object detection from point clouds , https://github.com/ open-mmlab/OpenPCDet, 2020

    work page 2020

  57. [57]

    V oxel- next: Fully sparse voxelnet for 3d object detection and tracking,

    Y . Chen, J. Liu, X. Zhang, X. Qi, and J. Jia, “V oxel- next: Fully sparse voxelnet for 3d object detection and tracking,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 21 674–21 683. Muhammad Shahbaz is passionate about in- terdisciplinary research across Advanced Computer Vision and AI and their applica...

    work page 2023

  58. [2025]

    [Online]

    DOI: 10.7910/DVN/D9SSWD. [Online]. Avail- able: https://doi.org/10.7910/DVN/D9SSWD

    work page doi:10.7910/dvn/d9sswd

This paper was first reviewed by deepseek-v4-flash on August 5, 2026.