Pith. sign in

REVIEW 4 major objections 7 minor 145 references

Towards Autonomous Riding: A Review of Perception, Planning, and Control in Intelligent Two-Wheelers

T0 review · 4 major / 7 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read This review argues that autonomous riding for two-wheelers is structurally behind autonomous driving, and that the missing pieces are large multimodal datasets, lightweight edge perception, and balance control.

desk verdict Useful first survey of autonomous riding for two-wheelers, but the dataset table in Section VI needs a verification pass before the 'no large 3D datasets' claim can be trusted. read the letter →

arxiv 2507.11852 v1 pith:5LL7DDRF submitted 2025-07-16 cs.RO cs.CV

classification cs.ROcs.CV
keywords autonomousridingmicromobilitye-scooterse-bikescomputervisiondeeplearningedgeAIdriving
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 review makes the case that autonomous riding for two-wheeled micromobility, e-scooters and e-bikes in particular, is significantly less mature than autonomous driving, and that the gap is structural rather than incidental. It argues that the car-first field has produced mature perception, planning, and control stacks, while two-wheeler research is confined to a handful of tasks such as helmet detection, rider detection, and violation spotting, with almost no work on 3D perception, trajectory forecasting, or balance control. The paper claims to be the first survey of autonomous riding from the computer vision, sensing, and deep learning perspectives, and that claim is the basis of the survey's value: if the selection is representative, the field gains a taxonomy, a gap list, and a technology-transfer roadmap from cars to two-wheelers.

What carries the argument

The organizing device is the autonomous driving perception–planning–control pipeline, applied as a taxonomy to two-wheeler literature. Figure 2 maps concrete tasks to each module, with solid arrows marking covered areas and dashed arrows marking modules the authors say remain completely unexplored, such as deeper behaviour analysis and certain manoeuvres. The companion tables compare sensor modalities, functional components, and datasets between autonomous driving and micromobility, and that comparison structure, rather than any single algorithm, is what carries the argument that AR lags AD.

What would settle it

A systematic literature search with a documented protocol would settle the question of representativeness: if it surfaces substantially more published two-wheeler perception, planning, or control work than the roughly 60 papers surveyed here, the paper's central gap claims weaken. Similarly, if deployed scooter-sharing fleets already collect synchronized multimodal sensor streams with long temporal sequences, the asserted absence of large-scale datasets would be incomplete.

Watch

Extended reading notes

Core claim

In the paper's own terms, the central discovery is that autonomous riding is an underexplored field whose core modules, perception, planning, and control, each lag their autonomous driving counterparts, and that the biggest bottleneck is data: micromobility datasets are small, 2D-only, and task-specific, whereas autonomous driving has large multimodal 3D benchmarks covering detection, segmentation, tracking, and motion forecasting. The authors show that existing two-wheeler perception work concentrates on a few tasks, mostly helmet and rider detection with YOLO-style detectors, while planning is limited to a small set of route-following e-scooter prototypes and control to a handful of balance experiments, one of which uses deep reinforcement learning to balance a scooter with a humanoid robot. From this they conclude that the way to accelerate autonomous riding is to transfer AD advances in sensor fusion, transformer-based perception, physics-informed prediction, and simulation while respecting the power, size, and cost constraints of two-wheelers.

Load-bearing premise

The review's gap list assumes that the roughly 60 papers it covers fairly represent the state of the art in autonomous riding, so important omitted work would change the reported gaps.

Editorial extensions

If this is right

  • The taxonomy gives the field a shared vocabulary: perception, planning, and control for two-wheelers, with specific tasks under each, so future papers can position themselves against a named gap.
  • If the data bottleneck is real, the priority should shift toward building large-scale multimodal micromobility datasets, including LiDAR, stereo, IMU, and long video sequences, before chasing new model architectures.
  • The identified transfer paths point to concrete work: adapting AD sensor fusion, transformer-based detection, and physics-informed trajectory prediction to the power and cost limits of e-scooters and e-bikes.
  • The control section singles out balance as the problem unique to two-wheelers and suggests reinforcement learning and physics-informed networks as the tools most likely to crack it.

Reading between the lines

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

  • A reader could push further than the paper does: if the data gap is the binding constraint, then synthetic augmentation through existing driving simulators extended with two-wheeler models may be the fastest route to progress, something the paper mentions briefly but does not develop.
  • The survey's AD-centric frame leans on perception and planning, but a safer two-wheeler may need human-machine interaction research as much as computer vision, since rider posture, gaze, and intent are part of the control loop; the paper touches this only in passing.
  • A natural testable extension of the gap list would be to build a standardized benchmark for e-scooter perception, with agreed train/test splits and metrics, the kind of artifact the paper calls for but does not itself provide.
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 / 7 minor

Summary. This paper is a survey of autonomous riding (AR) for two-wheeled vehicles—e-scooters, e-bikes, and motorcycles—organized around a perception-planning-control pipeline borrowed from autonomous driving (AD). The authors claim to provide the first comprehensive overview of micromobility methods from computer vision, sensing, and deep learning perspectives. The survey covers perception tasks (rider detection and tracking, re-identification and counting, violation and anomaly detection), planning (rider activity understanding, indoor and outdoor navigation, collision avoidance), and control (balance stabilization), and it compares each area with the state of the art in AD. It also presents a dataset comparison (Tables V and VI), discusses ethical and legal issues, and lists future directions such as multimodal lightweight sensors and edge deep learning. The paper's central diagnostic conclusion is that AR lags AD, particularly in the absence of large-scale 3D multimodal datasets, and that this gap is visible from the dataset tables.

Significance. If reliable, the survey would serve as a useful entry point for researchers and practitioners in micromobility, a field that has received less review attention than autonomous driving. The paper covers a set of relevant recent papers on e-scooter and motorcycle perception, naturalistic riding studies, and balance control, and it draws plausible connections to AD methods. The taxonomy and the explicit 'gaps and directions' sections are helpful for orienting newcomers. However, the current version's value is limited by the lack of a documented selection methodology and by citation and table errors that undermine the specific quantitative gap claims. The paper does not provide machine-checked proofs, reproducible code, or parameter-free derivations; its contribution is a narrative synthesis, and that synthesis is not yet trustworthy enough to support the headline claims.

major comments (4)
  1. [I (Contributions) and overall] The paper's central claim to be 'the first comprehensive overview of micromobility methods from the computer vision, sensing, and deep learning perspectives' is not supported by any documented selection methodology. The abstract and Section I assert a 'comprehensive' and 'systematic' review, but the paper never states the databases searched, the query terms, the inclusion/exclusion criteria, or the time window. As a result, claims such as 'the number of papers published within the last three years does not exceed ten' (Section III.A) and the assertion that the survey covers 'almost 60 papers' (Section VIII) cannot be checked, and the gap analysis may merely reflect the sample. I recommend adding a methods subsection that describes the search and screening process, and tempering the 'comprehensive' claim accordingly.
  2. [VI, Table VI] The dataset table is internally inconsistent and undermines the paper's main diagnostic conclusion, which explicitly asks readers to 'Examine Tables VI and V' to see the gaps. Specifically, the iRider [23] row reports a size of 21,454 and labels the data as 3D, but reference [23] is a four-page IEEE APSCON paper on biomechanical analysis; 21,454 exactly equals the size reported for the Apurv et al. [32] classification dataset in the same table, so this value appears to be a duplication error, and the '3D' label is unsupported by the cited paper. The Sabri et al. [38] entry lists 'Videos, 105,' yet the cited arXiv paper describes a much larger video collection, and the DashCop [74] and Gilroy et al. [37] entries also contain sensor/size values not clearly traceable to the cited sources. Because this table is the only quantitative evidence for the conclusion that no large-scale 3D micromobility datasets exist, a full verification pass against the cited papers is required before the conclusion can be accepted.
  3. [Reference list and in-text citations] Several duplicate and mismatched references prevent readers from verifying the survey's coverage. References [24] and [117] are the same paper (Poojari, Lee, and Paley, 'Outdoor localization and path planning for repositioning an autonomous electric scooter'); references [135] and [138] are both the Waymo Open Dataset paper. In Section III.A, the sentence 'Using edge AI platforms... [31] proposed to detect motorbikes in live video feeds' cites reference [31], which is the MobileNets paper (Howard, 2017), not a motorbike-detection work. In Section VI, the text says 'KITTI [140]', but reference [140] is the A*3D dataset paper, not KITTI. These errors are not merely cosmetic; they make it impossible to trust the reference list as the basis for a comprehensiveness claim.
  4. [VI, fourth paragraph] The strong negative claim that 'no large-scale datasets exist for 3D detection, 3D segmentation, temporal tracking, or rider behaviour prediction' is an overreach given the evidence presented. The table contains only eight micromobility datasets, and the paper provides no evidence that these eight are an exhaustive or systematic sample of the literature. Moreover, the 'Sensors' column is inconsistent: entries such as 'None' for Gilroy et al. [37] conflict with the cited paper's use of mobile phones/cameras, and the iRider [23] sensor list (IMU, GPS, cameras) is contradicted by the data size issue noted above. The claim should be rephrased as 'available datasets known to us' or supported by a systematic search and a consistently defined table schema.
minor comments (7)
  1. [V.A] The paragraph beginning '[130] suggested a system that allows electric scooters to balance while riding' is duplicated almost verbatim; please merge the two copies into one paragraph.
  2. [III.A] The citation bracket '[47], [47]' contains the same reference twice; this should be a single citation.
  3. [VI, second paragraph] The sentence 'The minimum number of datasets dedicated to optical flow exceeds ten' is confusing; it likely should say 'the number of datasets dedicated to optical flow exceeds ten,' and the comparison to all micromobility datasets should be stated more clearly.
  4. [Figure 2] The 'Sensors' item appears under the Planning module, but sensors are not a planning task; either move it or re-label the taxonomy to avoid confusion between sensing modalities and planning functions.
  5. [Table V] Several rows (e.g., Multi-Object Tracking with '34+' and Lane Detection with '10+') do not cite any source for the dataset counts; please add references or a note explaining how these counts were obtained.
  6. [I, Contributions] The contribution 'Surveying all the methods of micromobility riding' is not supported by the rest of the text, which states in Section VIII that 'almost 60 papers' were summarized; rephrase to reflect the actual scope.
  7. [II.C] The statement that KITTI and nuScenes 'include up to 12% two-wheeler instances' is attributed to the dataset papers [28], [29], but these papers do not appear to state this statistic; please provide a separate source or remove the specific number.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the survey synthesizes external literature; the few self-citations are peripheral and no derivation reduces to its own inputs.

full rationale

This paper is a literature review rather than a derivation or prediction pipeline, so the standard circularity patterns do not apply: there is no fitted parameter renamed as a prediction, no equation whose output is its input by construction, and no uniqueness theorem imported from the authors' prior work. The central comparative claim that autonomous riding lags autonomous driving in perception, planning, control, and datasets is supported by an enumeration of external references and by summary tables, not by an argument whose conclusion is assumed. The self-citations that do appear, [11] and [109], are not load-bearing: [11] is used as an in-text pointer near the 'first survey' claim and [109] is one Transformer baseline among many in a list of deep-learning methods. At most, the novelty claim might be weakly supported, but weak support is not circularity under the given standards. The potential inconsistencies in Table VI (e.g., the iRider dataset size matching the Apurv et al. entry) are accuracy and verification concerns about how faithfully external sources are reported, not evidence that a result is equivalent to its own input. The survey's organization, gap analysis, and technology-transfer roadmap are self-contained against external cited work, so the honest finding is no significant circularity.

Assumptions & free parameters 0 free parameters · 2 assumptions · 0 invented entities

The paper introduces no free parameters and no invented entities. Its conclusions depend on the representativeness of the cited corpus and on the choice of autonomous driving as the organizing framework; these are listed as domain assumptions.

assumptions (2)
  • domain assumption The surveyed corpus of roughly 60 papers is representative of the state of the art in autonomous riding.
    The review draws conclusions about gaps and progress from this corpus without a systematic search or inclusion protocol (Sections III through V).
  • domain assumption Autonomous driving is the appropriate reference framework for organizing autonomous riding tasks.
    The entire taxonomy maps AR onto the AD perception-planning-control pipeline (Sections I through III), which may not capture rider-specific dynamics.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards Autonomous Riding: A Review of Perception, Planning, and Control in Intelligent Two-Wheelers." pith.science (2026). https://pith.science/paper/5LL7DDRF

@misc{pith2026250711852,
  author       = {Pith},
  title        = {Pith review of: Towards Autonomous Riding: A Review of Perception, Planning, and Control in Intelligent Two-Wheelers},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5LL7DDRF}},
  note         = {Machine review of arXiv:2507.11852}
}
read the original abstract

The rapid adoption of micromobility solutions, particularly two-wheeled vehicles like e-scooters and e-bikes, has created an urgent need for reliable autonomous riding (AR) technologies. While autonomous driving (AD) systems have matured significantly, AR presents unique challenges due to the inherent instability of two-wheeled platforms, limited size, limited power, and unpredictable environments, which pose very serious concerns about road users' safety. This review provides a comprehensive analysis of AR systems by systematically examining their core components, perception, planning, and control, through the lens of AD technologies. We identify critical gaps in current AR research, including a lack of comprehensive perception systems for various AR tasks, limited industry and government support for such developments, and insufficient attention from the research community. The review analyses the gaps of AR from the perspective of AD to highlight promising research directions, such as multimodal sensor techniques for lightweight platforms and edge deep learning architectures. By synthesising insights from AD research with the specific requirements of AR, this review aims to accelerate the development of safe, efficient, and scalable autonomous riding systems for future urban mobility.

Figures

Figures reproduced from arXiv: 2507.11852 by the authors.

Figure 1
Figure 1. Visual explanation of the key safety challenges of [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The taxonomy of this survey, divided into three principal modules, Perception, Planning, and Control, and their key tasks. [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Comparative System Architectures for Autonomous Vehicles (a) Autonomous Driving System: Traditional vehicle [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Comparative analysis of sensing modalities for micro [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

145 extracted references · 60 canonical work pages

  1. [11]

    Visual attention methods in deep learning: An in-depth survey,

    M. Hassanin, S. Anwar, I. Radwan, F. S. Khan, and A. Mian, “Visual attention methods in deep learning: An in-depth survey,”Information Fusion, vol. 108, p. 102417, 2024

  2. [23]

    irider: Integrating sensors and cameras for in-depth biomechanical analysis of electric scooter,

    J. Alwin, S. J. Callista, K. Sharon, P. M. Kallarackal, R. Sanjai, J. M. Asensio-Gil, and C. R.-M. Garc ´ıa, “irider: Integrating sensors and cameras for in-depth biomechanical analysis of electric scooter,” in 2024 IEEE Applied Sensing Conference (APSCON). IEEE, 2024, pp. 1–4

  3. [32]

    Detection of E-scooter Riders in Naturalistic Scenes

    K. Apurv, R. Tian, and R. Sherony, “Detection of e-scooter riders in naturalistic scenes,”arXiv preprint arXiv:2111.14060, 2021

  4. [38]

    Detection of Micromobility Vehicles in Urban Traffic Videos

    K. Sabri, C. Djilali, G.-A. Bilodeau, N. Saunier, and W. Bouachir, “Detection of micromobility vehicles in urban traffic videos,”arXiv preprint arXiv:2402.18503, 2024

  5. [24]

    Outdoor localization and path planning for repositioning an autonomous electric scooter,

    S. S. Poojari, J. Lee, and D. A. Paley, “Outdoor localization and path planning for repositioning an autonomous electric scooter,”IEEE Transactions on Intelligent Vehicles, 2024

  6. [117]

    Outdoor localization and path planning for repositioning an autonomous electric scooter,

    S. S. Poojari, J. Lee, and D. A. Paley, “Outdoor localization and path planning for repositioning an autonomous electric scooter,”IEEE Transactions on Intelligent Vehicles, pp. 1–9, 2024

  7. [74]

    Dashcop: Automated e-ticket generation for two-wheeler traffic violations using dashcam videos,

    D. Rawat, K. Gupta, A. B. Roy, and R. K. Sarvadevabhatla, “Dashcop: Automated e-ticket generation for two-wheeler traffic violations using dashcam videos,” in2025 IEEE/CVF Winter Conference on Applica- tions of Computer Vision (WACV). IEEE, 2025, pp. 5387–5397

  8. [37]

    E-scooter rider detection and classification in dense urban environments,

    S. Gilroy, D. Mullins, E. Jones, A. Parsi, and M. Glavin, “E-scooter rider detection and classification in dense urban environments,”Results in engineering, vol. 16, p. 100677, 2022

  9. [138]

    Scalability in perception for autonomous driving: Waymo open dataset,

    P. Sun, H. Kretzschmar, X. Dotiwalla, A. Chouard, V . Patnaik, P. Tsui, J. Guo, Y . Zhou, Y . Chai, B. Caineet al., “Scalability in perception for autonomous driving: Waymo open dataset,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 2446–2454

  10. [31]

    Mobilenets: Efficient convolutional neural networks for mobile vision applications,

    A. G. Howard, “Mobilenets: Efficient convolutional neural networks for mobile vision applications,”arXiv preprint arXiv:1704.04861, 2017

  11. [140]

    A* 3d dataset: Towards autonomous driving in challenging environments,

    Q.-H. Pham, P. Sevestre, R. S. Pahwa, H. Zhan, C. H. Pang, Y . Chen, A. Mustafa, V . Chandrasekhar, and J. Lin, “A* 3d dataset: Towards autonomous driving in challenging environments,” in2020 IEEE In- ternational conference on Robotics and Automation (ICRA). IEEE, 2020, pp. 2267–2273

Show all 145 references
  1. [1]

    Micromobility,

    iMOVE Australia, “Micromobility,” https://imoveaustralia.com/topics/ micromobility/, 2024, accessed: 2025-05-05

  2. [2]

    Micro-mobility and road safety: Why do e-scooter riders use the sidewalk? evidence from a german field study,

    J. Anke, M. Ringhand, T. Petzoldt, and T. Gehlert, “Micro-mobility and road safety: Why do e-scooter riders use the sidewalk? evidence from a german field study,”European Transport Research Review, vol. 15, no. 1, p. 29, 2023

  3. [3]

    Autonomous driving cars in smart cities: Recent advances, requirements, and challenges,

    I. Yaqoob, L. U. Khan, S. A. Kazmi, M. Imran, N. Guizani, and C. S. Hong, “Autonomous driving cars in smart cities: Recent advances, requirements, and challenges,”IEEE Network, vol. 34, no. 1, pp. 174– 181, 2019

  4. [4]

    Shared autonomous micro-mobility for walkable cities,

    N. C. Sanchez and K. Larson, “Shared autonomous micro-mobility for walkable cities,”Transportation Research Interdisciplinary Perspec- tives, vol. 27, p. 101236, 2024

  5. [5]

    Building the computing system for autonomous micromobility vehicles: Design constraints and architectural optimizations,

    B. Yu, W. Hu, L. Xu, J. Tang, S. Liu, and Y . Zhu, “Building the computing system for autonomous micromobility vehicles: Design constraints and architectural optimizations,” in2020 53rd Annual IEEE/ACM International Symposium on Microarchitecture (MICRO). IEEE, 2020, pp. 1067–1081

  6. [6]

    Active safety systems for powered two-wheelers: A systematic review,

    G. Savino, R. Lot, M. Massaro, M. Rizzi, I. Symeonidis, S. Will, and J. Brown, “Active safety systems for powered two-wheelers: A systematic review,”Traffic injury prevention, vol. 21, no. 1, pp. 78–86, 2020. 14

  7. [7]

    Anti lock braking and vehicle stability control for motorcycles-why or why not,

    J. Gail, J. Funke, P. Seiniger, and U. Westerkamp, “Anti lock braking and vehicle stability control for motorcycles-why or why not,” in21st International Conference on the Enhanced Safety of Vehicles (ESV), Stuttgart, Germany, 2009

  8. [8]

    The future of the autonomous emergency braking for powered-two-wheelers: field testing end-users’ acceptability in realistic riding manoeuvres,

    M. Marra, C. Lucci, P. Huertas-Leyva, N. Baldanzini, M. Pierini, and G. Savino, “The future of the autonomous emergency braking for powered-two-wheelers: field testing end-users’ acceptability in realistic riding manoeuvres,” inIOP Conference Series: Materials Science and Engi...

  9. [9]

    Traction control of wheeled vehicles using dynamic feedback approach,

    N. Sarkar and X. Yun, “Traction control of wheeled vehicles using dynamic feedback approach,” inProceedings. 1998 IEEE/RSJ Inter- national Conference on Intelligent Robots and Systems. Innovations in Theory, Practice and Applications (Cat. No. 98CH36190), vol. 1. IEEE, 1998, p...

  10. [10]

    Dynamic supporting wheels for two-wheeler stability,

    V . Deepan, P. Jeyakumar, and S. Sreenath, “Dynamic supporting wheels for two-wheeler stability,” inAdvances in Design and Thermal Systems: Select Proceedings of ETDMMT 2020. Springer, 2021, pp. 67–79

  11. [12]

    Multi-sensor fusion and cooperative perception for autonomous driv- ing: A review,

    C. Xiang, C. Feng, X. Xie, B. Shi, H. Lu, Y . Lv, M. Yang, and Z. Niu, “Multi-sensor fusion and cooperative perception for autonomous driv- ing: A review,”IEEE Intelligent Transportation Systems Magazine, vol. 15, no. 5, pp. 36–58, 2023

  12. [13]

    A survey of motion planning and control techniques for self-driving urban vehicles,

    B. Paden, M. ˇC´ap, S. Z. Yong, D. Yershov, and E. Frazzoli, “A survey of motion planning and control techniques for self-driving urban vehicles,” IEEE Transactions on intelligent vehicles, vol. 1, no. 1, pp. 33–55, 2016

  13. [14]

    Deep rein- forcement learning framework for autonomous driving,

    A. E. Sallab, M. Abdou, E. Perot, and S. Yogamani, “Deep rein- forcement learning framework for autonomous driving,”arXiv preprint arXiv:1704.02532, 2017

  14. [15]

    Computer vision for autonomous vehicles: Problems, datasets and state-of-the-art,

    J. Janai, F. G ¨uney, A. Behl, and A. Geiger, “Computer vision for autonomous vehicles: Problems, datasets and state-of-the-art,”Foun- dations and Trends in Computer Graphics and Vision, vol. 12, no. 1-3, pp. 1–308, 2020

  15. [16]

    Using attention mechanisms in compact cnn models for improved micromobility safety through lane recognition

    C. Kaundanya, P. Cesar, B. Cronin, A. Fleury, M. Liu, and S. Little, “Using attention mechanisms in compact cnn models for improved micromobility safety through lane recognition.” inVEHITS, 2024, pp. 88–98

  16. [17]

    Livox mid-40 lidar sensor,

    Livox Technology, “Livox mid-40 lidar sensor,” https://livoxtech.com/ mid-40-and-mid-100, 2025, 38.4° circular FOV , 100 k pts/s, 260 m range

  17. [18]

    Titterton, J

    D. Titterton, J. Weston, and J. L. Weston,Strapdown Inertial Naviga- tion Technology, 2nd ed. Stevenage, UK: Institution of Engineering and Technology, 2004

  18. [19]

    Real-time roadway obstacle detection for electric scooters using deep learning and multi-sensor fusion,

    Z. Zheng, A. Hosseini, D. Chen, O. Shoghli, and A. Heydarian, “Real-time roadway obstacle detection for electric scooters using deep learning and multi-sensor fusion,”arXiv preprint arXiv:2504.03171, 2025

  19. [20]

    A wearable data collection system for studying micro-level e- scooter behavior in naturalistic road environment,

    A. Prabu, D. Shen, R. Tian, S. Chien, L. Li, Y . Chen, and R. Sherony, “A wearable data collection system for studying micro-level e- scooter behavior in naturalistic road environment,”arXiv preprint arXiv:2212.11979, 2022

  20. [21]

    Enabling autonomous navigation for affordable scooters,

    K. Liu and R. Mulky, “Enabling autonomous navigation for affordable scooters,”Sensors, vol. 18, no. 6, p. 1829, 2018

  21. [22]

    Motorbike driving activity recognition using smartphone motion sensors,

    A. Raheel, A. Arsalan, S. H. Noorani, S. Khan, M. Ehatisham-Ul-Haq, and Z. Ali, “Motorbike driving activity recognition using smartphone motion sensors,” in2023 25th International Multitopic Conference (INMIC). IEEE, 2023, pp. 1–6

  22. [25]

    A smart e- scooter with embedded estimation of rear vehicle trajectories for rider protection,

    H. Alai, W. Jeon, L. Alexander, and R. Rajamani, “A smart e- scooter with embedded estimation of rear vehicle trajectories for rider protection,”Mechanical Systems and Signal Processing, vol. 222, p. 111786, 2025

  23. [26]

    Fe-det: An effective traffic object detection framework for fish-eye cameras,

    X. Luo, Z. Cui, and F. Su, “Fe-det: An effective traffic object detection framework for fish-eye cameras,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, June 2024, pp. 7091–7099

  24. [27]

    Rear vehicle tracking on a smart e-scooter,

    H. Alai, W. Jeon, L. Alexander, and R. Rajamani, “Rear vehicle tracking on a smart e-scooter,” in2023 American Control Conference (ACC). IEEE, 2023, pp. 1735–1740

  25. [28]

    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,” in2012 IEEE Conference on Computer Vision and Pattern Recognition. IEEE, 2012, pp. 3354– 3361

  26. [29]

    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,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020, pp. 11...

  27. [30]

    Advances in electric two- wheeler technologies,

    A. K. Nayak, B. Ganguli, and P. M. Ajayan, “Advances in electric two- wheeler technologies,”Energy Reports, vol. 9, pp. 3508–3530, 2023

  28. [33]

    Mo- bilenetv2: Inverted residuals and linear bottlenecks,

    M. Sandler, A. Howard, M. Zhu, A. Zhmoginov, and L.-C. Chen, “Mo- bilenetv2: Inverted residuals and linear bottlenecks,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 4510–4520

  29. [34]

    Yolov3: an incremental improvement,

    J. Redmon, “Yolov3: an incremental improvement,” inComputer Vision and Pattern Recognition, 2018, p. 1

  30. [35]

    Yolov5-based electric scooter crackdown platform,

    S.-H. Lee, S.-H. Oh, and J.-G. Kim, “Yolov5-based electric scooter crackdown platform,”Applied Sciences, vol. 15, no. 6, p. 3112, 2025

  31. [36]

    Electric bicycle detection based on improved yolov5,

    C. Zhang, A. Xiong, X. Luo, C. Zhou, and J. Liang, “Electric bicycle detection based on improved yolov5,” in2022 4th International Conference on Advances in Computer Technology, Information Science and Communications (CTISC). IEEE, 2022, pp. 1–5

  32. [39]

    Development of an electric scooter photo recognition system using yolo

    C. Kim, S. Yu, and K. Y . Lee, “Development of an electric scooter photo recognition system using yolo.”Journal of Information Process- ing Systems, vol. 20, no. 6, 2024

  33. [40]

    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,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 779– 788

  34. [41]

    Performance evaluation of real-time ob- ject detection for electric scooters,

    D. Chen, A. Hosseini, A. Smith, A. F. Nikkhah, A. Heydarian, O. Shoghli, and B. Campbell, “Performance evaluation of real-time ob- ject detection for electric scooters,”arXiv preprint arXiv:2405.03039, 2024

  35. [42]

    Vehicle detection and tracking using yolo and deepsort,

    M. A. B. Zuraimi and F. H. K. Zaman, “Vehicle detection and tracking using yolo and deepsort,” in2021 IEEE 11th IEEE Symposium on Computer Applications & Industrial Electronics (ISCAIE). IEEE, 2021, pp. 23–29

  36. [43]

    Leveraging large language models for enhancing autonomous vehicle perception,

    A. Karagounis, “Leveraging large language models for enhancing autonomous vehicle perception,”arXiv preprint arXiv:2412.20230, 2024

  37. [44]

    Vulnerable road users detection using v2x communications,

    J. J. Anaya, E. Talavera, D. Gim ´enez, N. G ´omez, F. Jim´enez, and J. E. Naranjo, “Vulnerable road users detection using v2x communications,” in2015 IEEE 18th international conference on intelligent transporta- tion systems. IEEE, 2015, pp. 107–112

  38. [45]

    Vulnerability of deep learning model based anomaly detection in vehicle network,

    Y . Wang, D. W. M. Chia, and Y . Ha, “Vulnerability of deep learning model based anomaly detection in vehicle network,” in2020 IEEE 63rd International Midwest Symposium on Circuits and Systems (MWSCAS). IEEE, 2020, pp. 293–296

  39. [46]

    Multilevel framework to detect and handle vehicle occlusion,

    W. Zhang, Q. J. Wu, X. Yang, and X. Fang, “Multilevel framework to detect and handle vehicle occlusion,”IEEE Transactions on Intelligent Transportation Systems, vol. 9, no. 1, pp. 161–174, 2008

  40. [47]

    Drone-vs-bird detection challenge at ieee avss2021,

    A. Coluccia, A. Fascista, A. Schumann, L. Sommer, A. Dimou, D. Zarpalas, F. C. Akyon, O. Eryuksel, K. A. Ozfuttu, S. O. Altinuc et al., “Drone-vs-bird detection challenge at ieee avss2021,” in2021 17th IEEE International Conference on Advanced Video and Signal Based Surveillan...

  41. [48]

    More: A large-scale motorcycle re-identification dataset,

    A. Figueiredo, J. Brayan, R. O. Reis, R. Prates, and W. R. Schwartz, “More: A large-scale motorcycle re-identification dataset,” inProceed- ings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2021, pp. 4034–4043

  42. [49]

    Motorbike counting in heavily crowded scenes,

    C. K. Huynh, T. K. Dang, and C. A. Nguyen, “Motorbike counting in heavily crowded scenes,” inFuture Data and Security Engineering: 15 8th International Conference, FDSE 2021, Virtual Event, November 24–26, 2021, Proceedings 8. Springer, 2021, pp. 175–194

  43. [50]

    Person re-identification: Past, present and future,

    L. Zheng, Y . Yang, and A. G. Hauptmann, “Person re-identification: Past, present and future,”arXiv preprint arXiv:1610.02984, 2016

  44. [51]

    Vehicle re-identification method based on efficient self-attention cnn-transformer and multi-task learning op- timization,

    Y . Wang, R. Li, and Y . Shao, “Vehicle re-identification method based on efficient self-attention cnn-transformer and multi-task learning op- timization,”Sensors, vol. 25, no. 10, p. 2977, 2025

  45. [52]

    Robust and scalable vehicle re-identification via self-supervision,

    P. Khorramshahi, V . Shenoy, and R. Chellappa, “Robust and scalable vehicle re-identification via self-supervision,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 5295–5304

  46. [53]

    Day-night cross-domain vehicle re-identification,

    H. Li, J. Chen, A. Zheng, Y . Wu, and Y . Luo, “Day-night cross-domain vehicle re-identification,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 12 626–12 635

  47. [54]

    Tanet: Text region attention learning for vehicle re-identification,

    W. Hu, H. Zhan, P. Shivakumara, U. Pal, and Y . Lu, “Tanet: Text region attention learning for vehicle re-identification,”Engineering Applications of Artificial Intelligence, vol. 133, p. 108448, 2024

  48. [55]

    Vision-language constraint graph representation learning for unsuper- vised vehicle re-identification,

    D. Wang, Q. Wang, Z. Tu, W. Min, X. Xiong, Y . Zhong, and D. Gai, “Vision-language constraint graph representation learning for unsuper- vised vehicle re-identification,”Expert Systems with Applications, vol. 255, p. 124495, 2024

  49. [56]

    Clip-driven view-aware prompt learning for unsupervised vehicle re-identification,

    J. Xu, Q. Wang, X. Xiong, D. Gai, R. Zhou, and D. Wang, “Clip-driven view-aware prompt learning for unsupervised vehicle re-identification,” inProceedings of the AAAI Conference on Artificial Intelligence, vol. 39, no. 8, 2025, pp. 8896–8904

  50. [57]

    Implementation of an improved multi-object detection, tracking, and counting for autonomous driving,

    A. Albouchi, S. Messaoud, S. Bouaafia, M. A. Hajjaji, and A. Mtibaa, “Implementation of an improved multi-object detection, tracking, and counting for autonomous driving,”Multimedia Tools and Applications, vol. 83, no. 18, pp. 53 467–53 495, 2024

  51. [58]

    Video-based vehicle counting framework,

    Z. Dai, H. Song, X. Wang, Y . Fang, X. Yun, Z. Zhang, and H. Li, “Video-based vehicle counting framework,”IEEE access, vol. 7, pp. 64 460–64 470, 2019

  52. [59]

    Extremely overlapping vehicle counting,

    R. Guerrero-G ´omez-Olmedo, B. Torre-Jim ´enez, R. L ´opez-Sastre, S. Maldonado-Basc ´on, and D. Onoro-Rubio, “Extremely overlapping vehicle counting,” inPattern Recognition and Image Analysis: 7th Iberian Conference, IbPRIA 2015, Santiago de Compostela, Spain, June 17-19, 201...

  53. [60]

    Vehicle counting based on vehicle detection and tracking from aerial videos,

    X. Xiang, M. Zhai, N. Lv, and A. El Saddik, “Vehicle counting based on vehicle detection and tracking from aerial videos,”Sensors, vol. 18, no. 8, p. 2560, 2018

  54. [61]

    Portable roadside sensors for vehicle counting, classification, and speed measurement,

    S. Taghvaeeyan and R. Rajamani, “Portable roadside sensors for vehicle counting, classification, and speed measurement,”IEEE Transactions on Intelligent Transportation Systems, vol. 15, no. 1, pp. 73–83, 2013

  55. [62]

    Learning transferable visual models from natural language supervision,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clarket al., “Learning transferable visual models from natural language supervision,” inInternational conference on machine learning. PmLR, 2021, pp. 8748–8763

  56. [63]

    Emerging properties in self-supervised vision transformers,

    M. Caron, H. Touvron, I. Misra, H. J ´egou, J. Mairal, P. Bojanowski, and A. Joulin, “Emerging properties in self-supervised vision transformers,” inProceedings of the IEEE/CVF international conference on computer vision, 2021, pp. 9650–9660

  57. [64]

    Ssd: Single shot multibox detector,

    W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. Reed, C.-Y . Fu, and A. C. Berg, “Ssd: Single shot multibox detector,” inComputer Vision– ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part I 14. Springer, 2016, pp. 21– 37

  58. [65]

    Fast r-cnn,

    R. Girshick, “Fast r-cnn,” inProceedings of the IEEE international conference on computer vision, 2015, pp. 1440–1448

  59. [66]

    Identifying bikers without helmets using deep learning models,

    M. I. Hossain, R. B. Muhib, and A. Chakrabarty, “Identifying bikers without helmets using deep learning models,” in2021 Digital Image Computing: Techniques and Applications (DICTA). IEEE, 2021, pp. 01–08

  60. [67]

    An overview of the tesseract ocr engine,

    R. Smith, “An overview of the tesseract ocr engine,” inNinth interna- tional conference on document analysis and recognition (ICDAR 2007), vol. 2. IEEE, 2007, pp. 629–633

  61. [68]

    Robust automatic motorcycle helmet violation detection for an intelligent transportation system,

    D. N.-N. Tran, L. H. Pham, H.-J. Jeon, H.-H. Nguyen, H.-M. Jeon, T. H.-P. Tran, and J. W. Jeon, “Robust automatic motorcycle helmet violation detection for an intelligent transportation system,” inPro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognit...

  62. [69]

    ultralytics/yolov5: v3. 0,

    G. Jocher, A. Stoken, J. Borovec, L. Changyu, A. Hogan, L. Diaconu, J. Poznanski, L. Yu, P. Rai, R. Ferridayet al., “ultralytics/yolov5: v3. 0,”Zenodo, 2020

  63. [70]

    Real-time motorbike detection: Ai on the edge perspective,

    A. Akhtar, R. Ahmed, M. H. Yousaf, and S. A. Velastin, “Real-time motorbike detection: Ai on the edge perspective,”Mathematics, vol. 12, no. 7, p. 1103, 2024

  64. [71]

    Helmet rule violation detection for motorcyclists using a custom tracking framework and advanced object detection techniques,

    V . H. Duong, Q. H. Tran, H. S. P. Nguyen, D. Q. Nguyen, and T. C. Nguyen, “Helmet rule violation detection for motorcyclists using a custom tracking framework and advanced object detection techniques,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Re...

  65. [72]

    Simple online and realtime tracking with a deep association metric,

    N. Wojke, A. Bewley, and D. Paulus, “Simple online and realtime tracking with a deep association metric,” in2017 IEEE international conference on image processing (ICIP). IEEE, 2017, pp. 3645–3649

  66. [73]

    A computer vision system for detecting motorcycle violations in pedestrian zones,

    N. Hern ´andez-D´ıaz, Y . C. Pe˜naloza, Y . Y . Rios, J. C. Martinez-Santos, and E. Puertas, “A computer vision system for detecting motorcycle violations in pedestrian zones,”Multimedia Tools and Applications, pp. 1–24, 2024

  67. [75]

    Detecting, tracking and counting motorcycle rider traffic violations on unconstrained roads,

    A. Goyal, D. Agarwal, A. Subramanian, C. Jawahar, R. K. Sarvadev- abhatla, and R. Saluja, “Detecting, tracking and counting motorcycle rider traffic violations on unconstrained roads,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, p...

  68. [76]

    A coarse-to-fine two-stage helmet detection method for motorcyclists,

    H. Zhang, Z. Cui, and F. Su, “A coarse-to-fine two-stage helmet detection method for motorcyclists,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 7066–7074

  69. [77]

    Internimage: Exploring large-scale vision foun- dation models with deformable convolutions,

    W. Wang, J. Dai, Z. Chen, Z. Huang, Z. Li, X. Zhu, X. Hu, T. Lu, L. Lu, H. Liet al., “Internimage: Exploring large-scale vision foun- dation models with deformable convolutions,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023, pp. 14 ...

  70. [78]

    Prb-fpn+: Video analytics for enforcing motorcycle helmet laws,

    B.-S. Wang, P.-Y . Chen, Y .-K. Hsieh, J.-W. Hsieh, M.-C. Chang, J. He, S.-Y . Teng, H. Yue, and Y .-C. Tseng, “Prb-fpn+: Video analytics for enforcing motorcycle helmet laws,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 5477–5485

  71. [79]

    Motorcyclist helmet violation detection framework by leveraging robust ensemble and augmentation methods,

    T. Van Luong, H. S. P. Nguyen, D. K. Dinh, V . H. Duong, D. H. S. V o, H. Vu, M. T. Hoang, and T. C. Nguyen, “Motorcyclist helmet violation detection framework by leveraging robust ensemble and augmentation methods,” inProceedings of the IEEE/CVF Conference on Computer Vision ...

  72. [80]

    Detrs with collaborative hybrid assignments training,

    Z. Zong, G. Song, and Y . Liu, “Detrs with collaborative hybrid assignments training,” inProceedings of the IEEE/CVF international conference on computer vision, 2023, pp. 6748–6758

  73. [81]

    Efficientdet: Scalable and efficient object detection,

    M. Tan, R. Pang, and Q. V . Le, “Efficientdet: Scalable and efficient object detection,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 10 781–10 790

  74. [82]

    Helmet detection of motobike riders in real-world scenarios,

    B.-i. Kim, B. C. Ko, I.-s. Jang, and K.-J. Kim, “Helmet detection of motobike riders in real-world scenarios,” in2024 IEEE International Conference on Consumer Electronics-Asia (ICCE-Asia). IEEE, 2024, pp. 1–4

  75. [83]

    Weighted boxes fusion: Ensembling boxes from different object detection models,

    R. Solovyev, W. Wang, and T. Gabruseva, “Weighted boxes fusion: Ensembling boxes from different object detection models,”Image and Vision Computing, vol. 107, p. 104117, 2021

  76. [84]

    Cuda: Curriculum of data augmentation for long-tailed recognition,

    S. Ahn, J. Ko, and S.-Y . Yun, “Cuda: Curriculum of data augmentation for long-tailed recognition,”arXiv preprint arXiv:2302.05499, 2023

  77. [85]

    Robust motorcycle helmet detection in real-world scenarios: Using co-detr and minority class enhancement,

    H. V o, S. Tran, D. M. Nguyen, T. Nguyen, T. Do, D.-D. Le, and T. D. Ngo, “Robust motorcycle helmet detection in real-world scenarios: Using co-detr and minority class enhancement,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. ...

  78. [86]

    Detection of the triple riding and speed violation on two-wheelers using deep learning algorithms,

    N. C. Mallela, R. V olety, and N. RK, “Detection of the triple riding and speed violation on two-wheelers using deep learning algorithms,” Multimedia Tools and Applications, vol. 80, no. 6, pp. 8175–8187, 2021

  79. [87]

    Two-wheeler vehicle traffic violations detection and automated ticketing for indian road scenario,

    R. S. Charran and R. K. Dubey, “Two-wheeler vehicle traffic violations detection and automated ticketing for indian road scenario,”IEEE transactions on intelligent transportation systems, vol. 23, no. 11, pp. 22 002–22 007, 2022

  80. [88]

    Detection of license plate numbers and identification of non-helmet riders using yolo v2 and ocr method,

    B. Srilekha, K. Kiran, and V . V . P. Padyala, “Detection of license plate numbers and identification of non-helmet riders using yolo v2 and ocr method,” in2022 International Conference on Electronics and Renewable Systems (ICEARS). IEEE, 2022, pp. 1539–1549

  81. [89]

    Loltv: A low light two-wheeler violation dataset with anomaly detection technique,

    S. Bose, M. H. Kolekar, S. Nawale, and D. Khut, “Loltv: A low light two-wheeler violation dataset with anomaly detection technique,”IEEE Access, vol. 11, pp. 124 951–124 961, 2023. 16

  82. [90]

    Man vs. computer: Benchmarking machine learning algorithms for traffic sign recogni- tion,

    J. Stallkamp, M. Schlipsing, J. Salmen, and C. Igel, “Man vs. computer: Benchmarking machine learning algorithms for traffic sign recogni- tion,”Neural networks, vol. 32, pp. 323–332, 2012

  83. [91]

    Real-time traffic sign recognition and autonomous vehicle control system using convolutional neural networks,

    G. K. N G, A. Kishore, A. J. Krishnaet al., “Real-time traffic sign recognition and autonomous vehicle control system using convolutional neural networks,”Multimedia Tools and Applications, pp. 1–36, 2025

  84. [92]

    Traffic sign classification using cnn and detection using faster-rcnn and yolov4,

    N. Youssouf, “Traffic sign classification using cnn and detection using faster-rcnn and yolov4,”Heliyon, vol. 8, no. 12, 2022

  85. [93]

    Computer vision-based helmet use registration for e-scooter riders– the impact of the mandatory helmet law in copenhagen,

    F. W. Siebert, C. Riis, K. H. Janstrup, H. Lin, and F. B. H ¨uttel, “Computer vision-based helmet use registration for e-scooter riders– the impact of the mandatory helmet law in copenhagen,”Journal of safety research, vol. 87, pp. 257–265, 2023

  86. [94]

    Understand- ing spatio-temporal heterogeneity of bike-sharing and scooter-sharing mobility,

    R. Zhu, X. Zhang, D. Kondor, P. Santi, and C. Ratti, “Understand- ing spatio-temporal heterogeneity of bike-sharing and scooter-sharing mobility,”Computers, Environment and Urban Systems, vol. 81, p. 101483, 2020

  87. [95]

    Analysis of experimental data on dynamics and behavior of e-scooter riders and applications to the impact of automated driving functions on urban road safety,

    P. Brunner, A. L ¨ocken, F. Denk, R. Kates, and W. Huber, “Analysis of experimental data on dynamics and behavior of e-scooter riders and applications to the impact of automated driving functions on urban road safety,” in2020 IEEE Intelligent Vehicles Symposium (IV). IEEE, 202...

  88. [96]

    Watch out! e-scooter coming through!: Multimodal sensing of mixed traffic use and conflicts through riders’ ego-centric views,

    H. N. Kegalle, D. Hettiachchi, J. Chan, M. Sanderson, and F. D. Salim, “Watch out! e-scooter coming through!: Multimodal sensing of mixed traffic use and conflicts through riders’ ego-centric views,”Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technol...

  89. [97]

    Remote kinematic analysis for mobility scooter riders leveraging edge ai,

    T.-D. Nguyen, C. Zhang, M. Gitbumrungsin, A. Raheja, and T. Chen, “Remote kinematic analysis for mobility scooter riders leveraging edge ai,” inProceedings of the AAAI Symposium Series, vol. 4, no. 1, 2024, pp. 314–318

  90. [98]

    Beyond” tam- ing electric scooters

    M. Tabatabaie, S. He, H. Wang, and K. G. Shin, “Beyond” tam- ing electric scooters”: Disentangling understandings of micromobility naturalistic riding,”Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, vol. 8, no. 3, pp. 1–24, 2024

  91. [99]

    What factors contribute to e-scooter crashes: A first look using a naturalistic riding approach,

    E. White, F. Guo, S. Han, M. Mollenhauer, A. Broaddus, T. Sweeney, S. Robinson, A. Novotny, and R. Buehler, “What factors contribute to e-scooter crashes: A first look using a naturalistic riding approach,” Journal of safety research, vol. 85, pp. 182–191, 2023

  92. [100]

    Naturalistic e-scooter maneuver recognition with federated contrastive rider interaction learning,

    M. Tabatabaie and S. He, “Naturalistic e-scooter maneuver recognition with federated contrastive rider interaction learning,”Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, vol. 6, no. 4, pp. 1–27, 2023

  93. [101]

    Advanced driver assistance systems (adas),

    M. M. Antony and R. Whenish, “Advanced driver assistance systems (adas),” inAutomotive Embedded Systems: Key Technologies, Innova- tions, and Applications. Springer, 2021, pp. 165–181

  94. [102]

    Comprehensive study of driver behavior monitoring systems using computer vision and machine learning techniques,

    F. Qu, N. Dang, B. Furht, and M. Nojoumian, “Comprehensive study of driver behavior monitoring systems using computer vision and machine learning techniques,”Journal of Big Data, vol. 11, no. 1, p. 32, 2024

  95. [103]

    A multi-modal driver fatigue and distraction assessment system,

    C. Craye, A. Rashwan, M. S. Kamel, and F. Karray, “A multi-modal driver fatigue and distraction assessment system,”International Journal of Intelligent Transportation Systems Research, vol. 14, pp. 173–194, 2016

  96. [104]

    Awake at the wheel: Enhancing automotive safety through eeg-based fatigue detection,

    G. Siddhad, S. Dey, P. P. Roy, and M. Iwamura, “Awake at the wheel: Enhancing automotive safety through eeg-based fatigue detection,” in International Conference on Pattern Recognition. Springer, 2025, pp. 340–353

  97. [105]

    A multimodal approach to estimating vigilance using eeg and forehead eog,

    W.-L. Zheng and B.-L. Lu, “A multimodal approach to estimating vigilance using eeg and forehead eog,”Journal of neural engineering, vol. 14, no. 2, p. 026017, 2017

  98. [106]

    Driver sleepiness detection from eeg and eog signals using gan and lstm networks,

    Y . Jiao, Y . Deng, Y . Luo, and B.-L. Lu, “Driver sleepiness detection from eeg and eog signals using gan and lstm networks,”Neurocomput- ing, vol. 408, pp. 100–111, 2020

  99. [107]

    Deep learning for detecting multi-level driver fatigue using physiological signals: A comprehensive approach,

    M. Peivandi, S. Z. Ardabili, S. Sheykhivand, and S. Danishvar, “Deep learning for detecting multi-level driver fatigue using physiological signals: A comprehensive approach,”Sensors, vol. 23, no. 19, p. 8171, 2023

  100. [108]

    Attention is all you need,

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” Advances in neural information processing systems, vol. 30, 2017

  101. [109]

    Crossformer: Cross spatio-temporal transformer for 3d human pose estimation,

    M. Hassanin, A. Khamiss, M. Bennamoun, F. Boussaid, and I. Radwan, “Crossformer: Cross spatio-temporal transformer for 3d human pose estimation,”arXiv preprint arXiv:2203.13387, 2022

  102. [110]

    Towards evaluating driver fatigue with robust deep learning models,

    K. Alparslan, Y . Alparslan, and M. Burlick, “Towards evaluating driver fatigue with robust deep learning models,”arXiv preprint arXiv:2007.08453, 2020

  103. [111]

    Driver fatigue detection using ppg signal, facial features, head postures with an lstm model,

    L. Yu, X. Yang, H. Wei, J. Liu, and B. Li, “Driver fatigue detection using ppg signal, facial features, head postures with an lstm model,” Heliyon, vol. 10, no. 21, 2024

  104. [112]

    Can steering wheel detect your driving fatigue?

    J. Lu, X. Zheng, L. Tang, T. Zhang, Q. Z. Sheng, C. Wang, J. Jin, S. Yu, and W. Zhou, “Can steering wheel detect your driving fatigue?”IEEE Transactions on Vehicular Technology, vol. 70, no. 6, pp. 5537–5550, 2021

  105. [113]

    Intention-convolution and hybrid-attention network for vehicle trajectory prediction,

    C. Li, Z. Liu, S. Lin, Y . Wang, and X. Zhao, “Intention-convolution and hybrid-attention network for vehicle trajectory prediction,”Expert Systems with Applications, vol. 236, p. 121412, 2024

  106. [114]

    Kinodynamic trajectory optimization and control for car-like robots,

    C. R ¨osmann, F. Hoffmann, and T. Bertram, “Kinodynamic trajectory optimization and control for car-like robots,” in2017 IEEE/RSJ Inter- national Conference on Intelligent Robots and Systems (IROS). IEEE, 2017, pp. 5681–5686

  107. [115]

    Veo micromobility,

    Veo, “Veo micromobility,” 2025, accessed: 2025-04-29. [Online]. Available: https://www.veoride.com/

  108. [116]

    Autonomous scooter navigation for people with mobility challenges,

    R. S. Mulky, S. Koganti, S. Shahi, and K. Liu, “Autonomous scooter navigation for people with mobility challenges,” in2018 IEEE Inter- national Conference on Cognitive Computing (ICCC). IEEE, 2018, pp. 87–90

  109. [118]

    Intent-guided trajectory prediction for e-scooter riders and bicyclists,

    Z. Zhang, Z. Ding, Y . Chen, S. Chien, L. Li, R. Sheroy, J. Domeyer, and R. Tian, “Intent-guided trajectory prediction for e-scooter riders and bicyclists,” in2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC). IEEE, 2024, pp. 272–277

  110. [119]

    Human motion trajectory prediction: A survey,

    A. Rudenko, L. Palmieri, M. Herman, K. M. Kitani, D. M. Gavrila, and K. O. Arras, “Human motion trajectory prediction: A survey,”The International Journal of Robotics Research, vol. 39, no. 8, pp. 895–935, 2020

  111. [120]

    Vehicle trajectory prediction using LSTMs with spatial-temporal attention mechanisms,

    L. Lin, W. Li, H. Bi, and L. Qin, “Vehicle trajectory prediction using LSTMs with spatial-temporal attention mechanisms,”IEEE Intelligent Transportation Systems Magazine, vol. 14, no. 2, pp. 197—-208, 2022

  112. [121]

    Lstm-based graph attention network for vehicle trajectory prediction,

    J. Wang, K. Liu, and H. Li, “Lstm-based graph attention network for vehicle trajectory prediction,”Computer Networks, vol. 248, p. 110477, 2024

  113. [122]

    Multi-scale graph-transformer network for trajectory prediction of the autonomous vehicles,

    D. Singh and R. Srivastava, “Multi-scale graph-transformer network for trajectory prediction of the autonomous vehicles,”Intelligent Service Robotics, vol. 15, no. 3, pp. 307–320, 2022

  114. [123]

    Vehicle trajectory prediction and generation using lstm models and gans,

    L. Rossi, A. Ajmar, M. Paolanti, and R. Pierdicca, “Vehicle trajectory prediction and generation using lstm models and gans,”Plos one, vol. 16, no. 7, p. e0253868, 2021

  115. [124]

    Spectral temporal graph neural network for trajectory prediction,

    D. Cao, J. Li, H. Ma, and M. Tomizuka, “Spectral temporal graph neural network for trajectory prediction,” in2021 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2021, pp. 1839–1845

  116. [125]

    Trajectory prediction for autonomous driving using a transformer network,

    Z. Li and H. Yu, “Trajectory prediction for autonomous driving using a transformer network,”arXiv preprint arXiv:2402.16501, 2024

  117. [126]

    A physics-informed transformer model for vehicle trajectory prediction on highways,

    M. Geng, J. Li, Y . Xia, and X. M. Chen, “A physics-informed transformer model for vehicle trajectory prediction on highways,” Transportation research part C: emerging technologies, vol. 154, p. 104272, 2023

  118. [127]

    A deep reinforcement learning algorithm to control a two-wheeled scooter with a humanoid robot,

    J. Baltes, G. Christmann, and S. Saeedvand, “A deep reinforcement learning algorithm to control a two-wheeled scooter with a humanoid robot,”Engineering Applications of Artificial Intelligence, vol. 126, p. 106941, 2023

  119. [128]

    Introducing isaac gym: Accelerated reinforcement learning for robotics,

    NVIDIA, “Introducing isaac gym: Accelerated reinforcement learning for robotics,” https://developer.nvidia.com/blog/ introducing-isaac-gym-rl-for-robotics/, June 2021, accessed: 2025-04- 29

  120. [129]

    Prox- imal policy optimization algorithms,

    J. Schulman, F. Wolski, P. Dhariwal, A. Radford, and O. Klimov, “Prox- imal policy optimization algorithms,”arXiv preprint arXiv:1707.06347, 2017

  121. [130]

    A control framework for autonomous e-scooters,

    R. Soloperto, P. Wenzelburger, D. Meister, D. Scheuble, V . S. Breidohr, and F. Allg ¨ower, “A control framework for autonomous e-scooters,” IFAC-PapersOnLine, vol. 54, no. 2, pp. 252–258, 2021

  122. [131]

    The cityscapes dataset for semantic urban scene understanding,

    M. Cordts, M. Omran, S. Ramos, T. Rehfeld, M. Enzweiler, R. Be- nenson, U. Franke, S. Roth, and B. Schiele, “The cityscapes dataset for semantic urban scene understanding,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 3213–3223

  123. [132]

    Bdd100k: A diverse driving dataset for heterogeneous multitask learning,

    F. Yu, H. Chen, X. Wang, W. Xian, Y . Chen, F. Liu, V . Madhavan, and T. Darrell, “Bdd100k: A diverse driving dataset for heterogeneous multitask learning,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 2636–2645. 17

  124. [133]

    Argoverse: 3d tracking and forecasting with rich maps,

    M.-F. Chang, J. Lambert, P. Sangkloy, J. Singh, S. Bak, A. Hartnett, D. Wang, P. Carr, S. Lucey, D. Ramananet al., “Argoverse: 3d tracking and forecasting with rich maps,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 8748–8757

  125. [134]

    Gated2depth: Real-time dense lidar from gated images,

    T. Gruber, F. Julca-Aguilar, M. Bijelic, and F. Heide, “Gated2depth: Real-time dense lidar from gated images,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2019, pp. 1506–1516

  126. [136]

    Toronto-3d: A large-scale mobile lidar dataset for semantic segmen- tation of urban roadways,

    W. Tan, N. Qin, L. Ma, Y . Li, J. Du, G. Cai, K. Yang, and J. Li, “Toronto-3d: A large-scale mobile lidar dataset for semantic segmen- tation of urban roadways,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops, 2020, pp. 202– 203

  127. [137]

    The apolloscape dataset for autonomous driving,

    X. Huang, X. Cheng, Q. Geng, B. Cao, D. Zhou, P. Wang, Y . Lin, and R. Yang, “The apolloscape dataset for autonomous driving,” in Proceedings of the IEEE conference on computer vision and pattern recognition workshops, 2018, pp. 954–960

  128. [139]

    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,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 11 621–11 631

  129. [141]

    Carla: An open urban driving simulator,

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

  130. [142]

    Airsim: High-fidelity visual and physical simulation for autonomous vehicles,

    S. Shah, D. Dey, C. Lovett, and A. Kapoor, “Airsim: High-fidelity visual and physical simulation for autonomous vehicles,” inField and Service Robotics: Results of the 11th International Conference. Springer, 2018, pp. 621–635

  131. [143]

    Self-driving vehi- cles—an ethical overview,

    S. O. Hansson, M.- ˚A. Belin, and B. Lundgren, “Self-driving vehi- cles—an ethical overview,”Philosophy & Technology, vol. 34, no. 4, pp. 1383–1408, 2021

  132. [144]

    (2021) Electric scooters: micro-mobility, macro risk? [Online]

    Swiss Re Institute. (2021) Electric scooters: micro-mobility, macro risk? [Online]. Available: https://www.swissre.com/institute/research/ sonar/sonar2021/electric-scooters.html

  133. [145]

    An investigative study on the privacy implications of mobile e-scooter rental apps,

    N. Vinayaga-Sureshkanth, R. Wijewickrama, A. Maiti, and M. Jadli- wala, “An investigative study on the privacy implications of mobile e-scooter rental apps,” inProceedings of the 15th ACM conference on Security and Privacy in Wireless and Mobile Networks, 2022, pp. 125–139

  134. [146]

    (2020) The great e-scooter hack

    eurekalert. (2020) The great e-scooter hack. [Online]. Available: https://www.eurekalert.org/news-releases/627418

Pith tools

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