Pith. sign in

REVIEW 3 major objections 2 minor 1 cited by

A ring resonator on thin-film lithium niobate boosts integrated acousto-optic beam steering to 26% efficiency and enables on-chip FMCW LiDAR.

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 ↗

T0 review · grok-4.5

2026-07-15 11:54 UTC pith:5WMZ6A2K

load-bearing objection The abstract claims a useful TFLN AOBS+ring result (26% efficiency, 18° FOV, EO-locked FMCW), but the supplied full text is a different paper (MicroVision dataset), so the device claims cannot be audited. the 3 major comments →

arxiv 2603.18191 v2 pith:5WMZ6A2K submitted 2026-03-18 physics.app-ph

Resonance-enhanced integrated acousto-optic beam steering

Yue Yu , Qixuan Lin , Shucheng Fang , Joseph G. Thomas , Yibing Zhou , Zichen Xi , Jun Ji , Yizheng Zhu
show 3 more authors
Linbo Shao Bingzhao Li Mo Li
This is my paper
classification physics.app-ph
keywords acousto-optic beam steeringthin-film lithium niobateoptical ring resonatorelectro-optic lockingFMCW LiDARintegrated photonicsbeam steering
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.

Optical beam steering is essential for free-space links, sensing, and imaging, but mechanical scanners are slow and bulky while many solid-state approaches need complex pixelated arrays. This paper shows that integrating acousto-optic beam steering with an optical ring resonator on thin-film lithium niobate raises steering efficiency to as high as 26% over an 18° field of view, using fixed-wavelength light and on-chip acoustic transducers. The same platform’s electro-optic response is used to lock the resonator to a chirped laser, so the device can also run frequency-modulated continuous-wave LiDAR. The result is a compact, co-integrated chip that combines continuous one-dimensional beam steering with ranging capability without separate bulky optics or multi-wavelength sources.

Core claim

Integrating acousto-optic beam steering with an on-chip optical ring resonator on thin-film lithium niobate yields resonance-enhanced beam-steering efficiency up to 26% and an 18° field of view; electro-optic control can further lock the resonator to a chirped laser, enabling FMCW LiDAR on the same platform.

What carries the argument

Resonance-enhanced integrated acousto-optic beam steering: a ring resonator lengthens the light–sound interaction so a larger fraction of optical power is diffracted into the steered free-space beam, while co-located electro-optic electrodes lock the cavity to a chirped laser frequency.

Load-bearing premise

The reported 26% efficiency and 18° field of view are assumed to reflect usable free-space beam quality under realistic drive power, mode purity, and sidelobe levels, and the electro-optic lock is assumed to stay stable enough under laser chirp for real FMCW ranging.

What would settle it

Measure free-space far-field beam profiles and absolute diffracted power versus acoustic drive, plus residual frequency error of the locked resonator under a full FMCW chirp; if efficiency falls well below 26% once sidelobes and insertion loss are counted, or if lock fails across the chirp, the central performance claims do not hold.

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

If this is right

  • Continuous one-dimensional solid-state beam steering becomes practical without mechanical scanners or multi-wavelength laser arrays.
  • A single TFLN chip can combine steering and FMCW ranging for compact LiDAR or free-space optical links.
  • Piezoelectric and electro-optic functions of lithium niobate can be co-designed rather than treated as separate modules.
  • Fixed-wavelength sources suffice, simplifying system architecture compared with wavelength-tuned phased arrays.

Where Pith is reading between the lines

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

  • If the efficiency gain scales with resonator Q, higher-Q designs or better acoustic mode matching could push diffracted power still higher before nonlinear or thermal limits appear.
  • The same co-integration approach could be extended to two-dimensional steering by adding a second acoustic axis or cascaded resonators.
  • Residual phase noise from imperfect electro-optic locking under chirp would directly limit FMCW range resolution, so the lock bandwidth is a natural next specification to publish.

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

3 major / 2 minor

Summary. The submission is titled and abstracted as a device paper on resonance-enhanced integrated acousto-optic beam steering (AOBS) on thin-film lithium niobate: an optical ring resonator is claimed to raise beam-steering efficiency to 26% with an 18° field of view, and integrated electro-optic control is claimed to lock the resonator to a chirped laser for FMCW LiDAR. The body supplied under that title is, however, a completely different manuscript—MicroVision, an open image dataset and object-detection benchmark for vulnerable road users and micromobility vehicles (pedestrians, cyclists, e-scooterists, stationary bicycles/e-scooters) recorded from a VRU perspective in Gothenburg, with YOLO11/Faster R-CNN/RF-DETR baselines. No TFLN devices, acoustic transducers, ring resonators, far-field patterns, power budgets, or EO locking data appear in the full text.

Significance. If the AOBS abstract claims were supported by a matching experimental manuscript, co-integrating piezoelectric AOBS with EO resonance locking on TFLN would be a meaningful step toward compact solid-state beam steering and multifunctional LiDAR. That significance cannot be assessed here: the load-bearing efficiency, FOV, beam quality, and lock-stability results are stated only in the abstract and are not present in the provided body. Separately, the MicroVision body that was actually supplied is a useful open resource for micromobility safety vision (state-aware rider vs. parked-vehicle labels, VRU-perspective scenes, full-year coverage, released data/weights), but it is not the paper under review.

major comments (3)
  1. Title/abstract vs. full text mismatch: paper_id 2603.18191 and the abstract describe resonance-enhanced AOBS on TFLN (26% efficiency, 18° FOV, EO lock for FMCW LiDAR). The full manuscript text is MicroVision (arXiv-style 2603.18192 content)—a CV dataset paper with no photonic, acoustic, or LiDAR device content. The central AOBS claims therefore have no methods, figures, spectra, error bars, or baselines against which they can be checked.
  2. Because the AOBS experimental support is absent, the reader’s load-bearing conditions cannot be audited: free-space beam quality (mode purity, sidelobes), insertion loss, acoustic drive power for the stated 26% steering efficiency, and EO lock stability under chirp for FMCW ranging. These are not optional extras; they are required to substantiate the abstract’s headline metrics and the multifunctional-platform claim.
  3. Until a manuscript body that matches the AOBS title and abstract is provided, no technical recommendation on soundness of the device results is possible. Review of the MicroVision content would be a separate assignment for a different paper and venue.
minor comments (2)
  1. Abstract alone is well written for an AOBS device paper but is insufficient for peer review of experimental claims.
  2. If the MicroVision manuscript was intended for review instead, it should be submitted under its own title/ID with matching abstract; the current packaging conflates two unrelated works.

Circularity Check

0 steps flagged

No circular derivation: abstract reports experimental device metrics; supplied full text is a different paper, so no load-bearing AOBS derivation chain is available to reduce.

full rationale

The target paper (2603.18191) is presented only via its abstract, which claims measured outcomes—resonance-enhanced beam-steering efficiency up to 26%, 18° FOV, and electro-optic locking of a ring resonator to a chirped laser for FMCW LiDAR—not a first-principles derivation whose outputs equal its inputs by construction. There are no equations, fitted parameters re-labeled as predictions, uniqueness theorems, or ansatz-via-self-citation steps to audit. The CACHEABLE full manuscript is instead MicroVision (2603.18192), a computer-vision dataset/benchmark paper with no TFLN, AOBS, resonators, or beam-steering content; that mismatch prevents any deeper walk of an AOBS derivation chain. On the available abstract content alone, nothing reduces by definition or by self-citation to its own inputs. Score 0 is therefore the honest finding: no significant circularity is identifiable.

Axiom & Free-Parameter Ledger

0 free parameters · 3 axioms · 0 invented entities

Abstract-only experimental device paper. Load-bearing content is measured performance and platform physics (piezoelectric SAW/BAW interaction with guided light; EO tuning of a ring), not free parameters of a theory fit. No new fundamental entities are postulated. Assumptions are standard domain physics of TFLN acousto-optics and electro-optics plus unstated measurement conventions for ‘efficiency’ and FOV.

axioms (3)
  • domain assumption Acoustic waves on TFLN can phase-match and diffract guided optical modes to produce continuous angular beam steering at fixed optical wavelength.
    Core AOBS operating principle assumed throughout the abstract.
  • domain assumption An optical ring resonator increases acousto-optic interaction length/strength enough to raise diffraction efficiency without destroying usable beam quality or FOV.
    Central design premise for ‘resonance-enhanced’ efficiency up to 26%.
  • domain assumption Integrated electro-optic electrodes can lock the ring resonance to a chirped laser frequency with sufficient bandwidth and stability for FMCW LiDAR.
    Required for the multifunctional LiDAR claim; not demonstrated in the abstract text alone.

pith-pipeline@v1.1.0-grok45 · 17863 in / 2508 out tokens · 27069 ms · 2026-07-15T11:54:41.322988+00:00 · methodology

0 comments
read the original abstract

Optical beam steering is a key technology for free-space optical communication, sensing, and imaging. Mechanical beam steering systems suffer from limited scanning speed and bulky form factors, while existing solid-state solutions rely on pixelated synthetic aperture that requires complex fabrication and control architectures. Integrated acousto-optic beam steering (AOBS) is an emerging technology that enables continuous one-dimensional beam steering using integrated acoustic transducers and fixed-wavelength laser sources. Here, we integrate AOBS with an optical ring resonator on the same thin-film lithium niobate (TFLN) platform to significantly enhance beam steering efficiency and system functionality. The resulting device achieves a resonance-enhanced beam steering efficiency of up to $26\%$ and a field of view of $18^\circ$. Moreover, by leveraging integrated electro-optic control, we dynamically lock the ring-resonator's resonance to a chirped laser frequency, enabling frequency-modulated continuous-wave (FMCW) LiDAR operation. By combining lithium niobate's piezoelectric and electro-optic properties, this work establishes a compact, efficient, and scalable beam-steering platform with co-integrated acousto-optic modulation and electro-optic control for multifunctional applications.

discussion (0)

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

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. 2D Optical Beam Scanning using Integrated Acousto-Optics and a Frequency Comb

    physics.optics 2026-05 unverdicted novelty 6.0

    Integrated acousto-optic deflection combined with frequency comb dispersion achieves 2D beam scanning from one aperture, with a demonstrated 18.2 by 4.3 degree field of view using 11 comb lines.

Reference graph

Works this paper leans on

43 extracted references · 15 canonical work pages · cited by 1 Pith paper

  1. [1]

    URLhttps://doi.org/10.4271/J3194_201911

    SAE International.Taxonomy and Classification of Powered Micromobility Vehicles, SAE Stan- dard J3194_201911, November 2019. URLhttps://doi.org/10.4271/J3194_201911

    work page doi:10.4271/j3194_201911 2019

  2. [2]

    A. G. Olabi, Tabbi Wilberforce, Khaled Obaideen, Enas Taha Sayed, Nabila Shehata, Abdul Hai Alami, and Mohammad Ali Abdelkareem. Micromobility: Progress, benefits, challenges, policy and regulations, energy sources and storage, and its role in achieving sustainable development goals.International Journal of Thermofluids, 17(January):100292, 2023. ISSN 266...

    doi:10.1016/j.ijft.2023.100292 2023

  3. [3]

    Transportation Transformation: Is Micromobility Making a Macro Impact on Sustainability?Journal of Planning Literature, 36(1):46–61, 2021

    Michael McQueen, Gabriella Abou-Zeid, John MacArthur, and Kelly Clifton. Transportation Transformation: Is Micromobility Making a Macro Impact on Sustainability?Journal of Planning Literature, 36(1):46–61, 2021. ISSN 15526593. doi:10.1177/0885412220972696

    work page doi:10.1177/0885412220972696 2021

  4. [4]

    Safer Micromobility: Tech- nical Background Report The International Transport Forum

    George Yannis, Virginia Petraki, and Philippe Crist. Safer Micromobility: Tech- nical Background Report The International Transport Forum. Technical Re- port March, 2024. URL https://www.itf-oecd.org/sites/default/files/ safer-micrombility-technical-report.pdf

    2024

  5. [5]

    Advancing traffic safety through the safe system approach: A systematic review.Accident Analysis & Prevention, 199(February):107518, may 2024

    Md Nasim Khan and Subasish Das. Advancing traffic safety through the safe system approach: A systematic review.Accident Analysis & Prevention, 199(February):107518, may 2024. ISSN 00014575. doi:10.1016/j.aap.2024.107518. URL https://doi.org/10.1016/j.aap.2024. 107518https://linkinghub.elsevier.com/retrieve/pii/S0001457524000630

    work page doi:10.1016/j.aap.2024.107518 2024

  6. [6]

    Marco Dozza, Alessio Violin, and Alexander Rasch. A data-driven framework for the safe integration of micro-mobility into the transport system: Comparing bicycles and e-scooters in field trials.Journal of Safety Research, 81:67–77, jun 2022. ISSN 00224375. doi:10.1016/j.jsr.2022.01.007. URL https://doi.org/10.1016/j.jsr.2022. 01.007https://linkinghub.else...

    work page doi:10.1016/j.jsr.2022.01.007 2022

  7. [7]

    How do different micro-mobility vehicles affect longitudinal control? Results from a field experiment.Journal of Safety Research, 84:24–32, feb 2023

    Marco Dozza, Tianyou Li, Lucas Billstein, Christoffer Svernlöv, and Alexander Rasch. How do different micro-mobility vehicles affect longitudinal control? Results from a field experiment.Journal of Safety Research, 84:24–32, feb 2023. ISSN 00224375. doi:10.1016/j.jsr.2022.10.005. URL https://doi.org/10.1016/j.jsr.2022. 10.005https://linkinghub.elsevier.co...

    work page doi:10.1016/j.jsr.2022.10.005 2023

  8. [8]

    E-Scooter Rider detection and classification in dense urban environments.Results in Engineering, 16 (October):100677, 2022

    Shane Gilroy, Darragh Mullins, Edward Jones, Ashkan Parsi, and Martin Glavin. E-Scooter Rider detection and classification in dense urban environments.Results in Engineering, 16 (October):100677, 2022. ISSN 25901230. doi:10.1016/j.rineng.2022.100677. URL https: //doi.org/10.1016/j.rineng.2022.100677

    doi:10.1016/j.rineng.2022.100677 2022

  9. [9]

    Electric scooter safety: An integrative review of evidence from transport and medical research do- mains.Sustainable Cities and Society, 89(November 2022):104313, feb 2023

    Khashayar Kazemzadeh, Milad Haghani, and Frances Sprei. Electric scooter safety: An integrative review of evidence from transport and medical research do- mains.Sustainable Cities and Society, 89(November 2022):104313, feb 2023. ISSN 22106707. doi:10.1016/j.scs.2022.104313. URL https://doi.org/10.1016/j.scs.2022. 104313https://linkinghub.elsevier.com/retr...

    work page doi:10.1016/j.scs.2022.104313 2022

  10. [10]

    Understanding factors influencing e-scooterist crash risk: A naturalistic study of rental e-scooters in an urban area.Accident Analysis and Prevention, 209(June 2024):107839, 2025

    Rahul Rajendra Pai and Marco Dozza. Understanding factors influencing e-scooterist crash risk: A naturalistic study of rental e-scooters in an urban area.Accident Analysis and Prevention, 209(June 2024):107839, 2025. ISSN 00014575. doi:10.1016/j.aap.2024.107839. URL https: //doi.org/10.1016/j.aap.2024.107839. 12

    work page doi:10.1016/j.aap.2024.107839 2024

  11. [11]

    E- Scooter Presence in Urban Areas: Are Consistent Rules, Paying Attention and Smooth In- frastructure Enough for Safety?Sustainability (Switzerland), 14(21), 2022

    Matteo della Mura, Serena Failla, Nicolò Gori, Alfonso Micucci, and Filippo Paganelli. E- Scooter Presence in Urban Areas: Are Consistent Rules, Paying Attention and Smooth In- frastructure Enough for Safety?Sustainability (Switzerland), 14(21), 2022. ISSN 20711050. doi:10.3390/su142114303

    work page doi:10.3390/su142114303 2022

  12. [12]

    Zenseact open dataset: A large-scale and diverse multimodal dataset for autonomous driving

    Mina Alibeigi, William Ljungbergh, Adam Tonderski, Georg Hess, Adam Lilja, Carl Lindstrom, Daria Motorniuk, Junsheng Fu, Jenny Widahl, and Christoffer Petersson. Zenseact open dataset: A large-scale and diverse multimodal dataset for autonomous driving. InProceedings of the IEEE/CVF International Conference on Computer Vision, 2023

    2023

  13. [13]

    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.The International Journal of Robotics Research, (October):1–6, 2013

    2013

  14. [14]

    Pei Sun, Henrik Kretzschmar, Xerxes Dotiwalla, Aurélien Chouard, Vijaysai Patnaik, Paul Tsui, James Guo, Yin Zhou, Yuning Chai, Benjamin Caine, Vijay Vasudevan, Wei Han, Jiquan Ngiam, Hang Zhao, Aleksei Timofeev, Scott Ettinger, Maxim Krivokon, Amy Gao, Aditya Joshi, Yu Zhang, Jonathon Shlens, Zhifeng Chen, and Dragomir Anguelov. Scalability in perception...

    doi:10.1109/cvpr42600.2020.00252 2020

  15. [15]

    Comparison of E-Scooter and Bike Users’ Behavior in Mixed Traffic.Transportation Research Record, 2024

    Natalia Distefano, Salvatore Leonardi, Mariusz Kie´c, and Carmelo D’Agostino. Comparison of E-Scooter and Bike Users’ Behavior in Mixed Traffic.Transportation Research Record, 2024. ISSN 21694052. doi:10.1177/03611981241263339

    work page doi:10.1177/03611981241263339 2024

  16. [16]

    García-Venegas, D

    M. García-Venegas, D. A. Mercado-Ravell, L. A. Pinedo-Sánchez, and C. A. Carballo-Monsivais. On the safety of vulnerable road users by cyclist detection and tracking.Machine Vision and Applications, 32(5):1–17, 2021. ISSN 14321769. doi:10.1007/s00138-021-01231-4. URL https://doi.org/10.1007/s00138-021-01231-4

    work page doi:10.1007/s00138-021-01231-4 2021

  17. [17]

    Xiaofei Li, Fabian Flohr, Yue Yang, Hui Xiong, Markus Braun, Shuyue Pan, Keqiang Li, and Dariu M. Gavrila. A new benchmark for vision-based cyclist detection. In2016 IEEE Intelligent Vehicles Symposium (IV), volume 2016-Augus, pages 1028–1033. IEEE, jun 2016. ISBN 978-1-5090-1821-5. doi:10.1109/IVS.2016.7535515. URL http://ieeexplore.ieee. org/document/7535515/

    work page doi:10.1109/ivs.2016.7535515 2016

  18. [18]

    Detection of e-scooter riders in naturalistic scenes, 2021

    Kumar Apurv, Renran Tian, and Rini Sherony. Detection of e-scooter riders in naturalistic scenes, 2021. URLhttps://arxiv.org/abs/2111.14060

    Pith/arXiv arXiv 2021

  19. [19]

    Performance Evaluation of Real-Time Object Detection for Electric Scooters, 2024

    Dong Chen, Arman Hosseini, Arik Smith, Amir Farzin Nikkhah, Arsalan Heydarian, Omid Shoghli, and Bradford Campbell. Performance Evaluation of Real-Time Object Detection for Electric Scooters, 2024. URLhttps://arxiv.org/abs/2405.03039

    Pith/arXiv arXiv 2024

  20. [20]

    Detection of Micromobility Vehicles in Urban Traffic Videos.Proceedings of the Conference on Robots and Vision, (Mmv), may 2024

    Khalil Sabri, Célia Djilali, Guillaume-Alexandre Bilodeau, Nicolas Saunier, and Wassim Bouachir. Detection of Micromobility Vehicles in Urban Traffic Videos.Proceedings of the Conference on Robots and Vision, (Mmv), may 2024. doi:10.21428/d82e957c.abc3243f. URL https://crv.pubpub.org/pub/du7cg0ee

    work page doi:10.21428/d82e957c.abc3243f 2024

  21. [21]

    Bot-sort: Robust associations multi- pedestrian tracking, 2022

    Nir Aharon, Roy Orfaig, and Ben-Zion Bobrovsky. Bot-sort: Robust associations multi- pedestrian tracking, 2022. URLhttps://arxiv.org/abs/2206.14651

    Pith/arXiv arXiv 2022

  22. [22]

    Pedestrians: 2019 data (Traffic Safety Facts

    National Center for Statistics and Analysis. Pedestrians: 2019 data (Traffic Safety Facts. Report No. DOT HS 813 079). Technical Report May, National Highway Traffic Safety Administration,

    2019

  23. [23]

    URLhttps://crashstats.nhtsa.dot.gov/Api/Public/Publication/813079

  24. [24]

    Lawrence Zitnick

    Tsung Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Dollár, and C. Lawrence Zitnick. Microsoft COCO: Common objects in context.Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 8693 LNCS(PART 5):740–755, 2014. ISSN 16113349. doi:10.1...

    doi:10.1007/978-3-319-10602-1_48 2014

  25. [25]

    Yolov7: Trainable bag-of- freebies sets new state-of-the-art for real-time object detectors, 2022

    Chien-Yao Wang, Alexey Bochkovskiy, and Hong-Yuan Mark Liao. Yolov7: Trainable bag-of- freebies sets new state-of-the-art for real-time object detectors, 2022. URL https://arxiv. org/abs/2207.02696

    Pith/arXiv arXiv 2022

  26. [26]

    Threat assessment from naturalistic video-data: How to detect, classify, and estimate the position of multiple road users from cameras, 2024

    Luhan Fang and Yahui Wu. Threat assessment from naturalistic video-data: How to detect, classify, and estimate the position of multiple road users from cameras, 2024. URL https: //odr.chalmers.se/items/1f59fb35-7b6a-49b6-958e-d0fe540917ee. 13

    2024

  27. [27]

    Label Studio: Data labeling software, 2025

    Maxim Tkachenko, Mikhail Malyuk, Andrey Holmanyuk, and Nikolai Liubimov. Label Studio: Data labeling software, 2025. URLhttps://github.com/HumanSignal/label-studio

    2025

  28. [28]

    The Hungarian method for the assignment problem.Naval Research Logistics Quarterly, 2(1-2):83–97, 1955

    H W Kuhn. The Hungarian method for the assignment problem.Naval Research Logistics Quarterly, 2(1-2):83–97, 1955. doi:https://doi.org/10.1002/nav.3800020109. URL https: //onlinelibrary.wiley.com/doi/abs/10.1002/nav.3800020109

    doi:10.1002/nav.3800020109 1955

  29. [29]

    Mark Everingham, Luc Van Gool, Christopher K. I. Williams, John Winn, and Andrew Zisser- man. The Pascal Visual Object Classes (VOC) Challenge.International Journal of Computer Vision, 88(2):303–338, jun 2010. ISSN 0920-5691. doi:10.1007/s11263-009-0275-4. URL http://link.springer.com/10.1007/s11263-009-0275-4

    doi:10.1007/s11263-009-0275-4 2010

  30. [30]

    A coefficient of agreement for nominal scales.Educational and Psy- chological Measurement, 20:37–46, 1960

    Jacob Cohen. A coefficient of agreement for nominal scales.Educational and Psy- chological Measurement, 20:37–46, 1960. ISSN 1552-3888(Electronic),0013-1644(Print). doi:10.1177/001316446002000104

    doi:10.1177/001316446002000104 1960

  31. [31]

    Ultralytics YOLO11, 2024

    Glenn Jocher and Jing Qiu. Ultralytics YOLO11, 2024. URL https://github.com/ ultralytics/ultralytics

    2024

  32. [32]

    Faster R-CNN: Towards Real- Time Object Detection with Region Proposal Networks

    Shaoqing Ren, Kaiming He, Ross Girshick, and Jian Sun. Faster R-CNN: Towards Real- Time Object Detection with Region Proposal Networks. In C Cortes, N Lawrence, D Lee, M Sugiyama, and R Garnett, editors,Advances in Neural Information Processing Systems, vol- ume 28. Curran Associates, Inc., 2015. URL https://proceedings.neurips.cc/paper_ files/paper/2015/...

    2015

  33. [33]

    RF- DETR: Neural Architecture Search for Real-Time Detection Transformers, 2025

    Isaac Robinson, Peter Robicheaux, Matvei Popov, Deva Ramanan, and Neehar Peri. RF- DETR: Neural Architecture Search for Real-Time Detection Transformers, 2025. URL https: //arxiv.org/abs/2511.09554

    arXiv 2025

  34. [34]

    Detectron2,

    Yuxin Wu, Alexander Kirillov, Francisco Massa, Wan-Yen Lo, and Ross Girshick. Detectron2,

  35. [35]

    URLhttps://github.com/facebookresearch/detectron2

  36. [36]

    Springer US, Boston, MA, 2009

    Steven M Beitzel, Eric C Jensen, and Ophir Frieder.MAP, pages 1691–1692. Springer US, Boston, MA, 2009. ISBN 978-0-387-39940-9. doi:10.1007/978-0-387-39940-9_492. URL https://doi.org/10.1007/978-0-387-39940-9_492

    work page doi:10.1007/978-0-387-39940-9_492 2009

  37. [37]

    Slicing Aided Hyper Inference and Fine-Tuning for Small Object Detection

    Fatih Cagatay Akyon, Sinan Onur Altinuc, and Alptekin Temizel. Slicing Aided Hyper Inference and Fine-Tuning for Small Object Detection. In2022 IEEE In- ternational Conference on Image Processing (ICIP), pages 966–970. IEEE, oct

  38. [38]

    doi:10.1109/ICIP46576.2022.9897990

    ISBN 978-1-6654-9620-9. doi:10.1109/ICIP46576.2022.9897990. URL http://arxiv.org/abs/2202.06934http://dx.doi.org/10.1109/ICIP46576. 2022.9897990https://ieeexplore.ieee.org/document/9897990/

    doi:10.1109/icip46576.2022.9897990 2022

  39. [39]

    Ben Beck, Derek Chong, Jake Olivier, Monica Perkins, Anthony Tsay, Adam Rushford, Lingxiao Li, Peter Cameron, Richard Fry, and Marilyn Johnson. How much space do drivers provide when passing cyclists? Understanding the impact of motor vehicle and infrastructure characteristics on passing distance.Accident Analysis & Prevention, 128: 253–260, jul 2019. ISS...

    work page doi:10.1016/j.aap.2019.03.007 2019

  40. [40]

    Marco Dozza and Julia Werneke. Introducing naturalistic cycling data: What factors influence bicyclists’ safety in the real world?Transportation Research Part F: Traffic Psychology and Behaviour, 24:83–91, may 2014. ISSN 1369-8478. doi:10.1016/J.TRF.2014.04.001. URL https://www.sciencedirect.com/science/article/pii/S1369847814000394

    work page doi:10.1016/j.trf.2014.04.001 2014

  41. [41]

    Schleinitz, T

    K. Schleinitz, T. Petzoldt, L. Franke-Bartholdt, J. Krems, and T. Gehlert. The German Naturalis- tic Cycling Study – Comparing cycling speed of riders of different e-bikes and conventional bicy- cles.Safety Science, 92:290–297, feb 2017. ISSN 0925-7535. doi:10.1016/J.SSCI.2015.07.027. URLhttps://www.sciencedirect.com/science/article/pii/S0925753515001976

    doi:10.1016/j.ssci.2015.07.027 2017

  42. [42]

    A Low-Cost Video-Based Solution for City- Wide Bicycle Counting in Starter Cities

    Eduardo Peixoto, João Moutinho, and Rui José. A Low-Cost Video-Based Solution for City- Wide Bicycle Counting in Starter Cities. In Henrique Santos, Gabriela Viale Pereira, Matthias Budde, Sérgio F Lopes, and Predrag Nikolic, editors,Science and Technologies for Smart Cities, pages 139–150, Cham, 2020. Springer International Publishing. ISBN 978-3-030-51005-3

    2020

  43. [43]

    AI-Enhanced Road Safety: Real-Time Information on Cycle Traffic on Rural Roads

    Francisco Vacalebri, Sara Moll, Griselda López, and Alfredo García. AI-Enhanced Road Safety: Real-Time Information on Cycle Traffic on Rural Roads. In Angel A Juan, Javier Faulin, and 14 David Lopez-Lopez, editors,Decision Sciences, pages 281–288, Cham, 2025. Springer Nature Switzerland. ISBN 978-3-031-78241-1. 15

    2025