Pith. sign in

REVIEW 5 major objections 6 minor 57 references

Multimodal Large Language Models for Enhanced Traffic Safety: A Comprehensive Review and Future Trends

T0 review · 5 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read This review argues that multimodal large language models improve traffic safety by fusing vision, language, and sensor data, reporting accuracy gains and stronger defenses against adversarial attacks.

desk verdict A survey whose quantitative case is built on mismatched citations and non-MLLM systems; useful only as a literature pointer until the evidence is redone. read the letter →

arxiv 2504.16134 v1 pith:IQJKPMZP submitted 2025-04-21 cs.CV cs.CL

Mohammad Abu Tami , Mohammed Elhenawy , Huthaifa I. Ashqar This is my paper · ORCID
classification cs.CVcs.CL
keywords multimodallargelanguagemodelstrafficsafetyadvanceddriver-assistancesystemsadversarialrobustnessautonomousdrivingsensorfusionedgedeploymentcontextualreasoning
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 sets out to establish that multimodal large language models (MLLMs), by integrating visual, linguistic, and environmental sensor data, can overcome the fragmented-perception and adversarial-vulnerability problems of conventional driver-assistance systems. It compares traditional ADAS with MLLM-based systems across accuracy, speed, energy, and robustness, reporting gains such as rainy-condition mIoU rising from 72.3% to 88.6% and adversarial misclassification falling from 89% to 12%. A sympathetic reader would care because these numbers point toward a concrete path from brittle single-sensor pipelines to context-aware, edge-deployable safety systems.

What carries the argument

The carrying mechanism is cross-modal fusion inside an LLM backbone: image, LiDAR, text, GPS, and weather signals are aligned into one representation so that scene understanding is contextual rather than per-sensor. Named systems do the work: LLaVA-ST aligns language and visual tokens in space and time for trajectory prediction; SeeUnsafe uses severity-based aggregation for interactive video analysis; AccidentGPT fuses accident reports with real-time sensor streams to flag collision-prone areas. The shared mechanism is that each modality can validate or disambiguate the others, which is what the paper credits for both accuracy gains and adversarial robustness.

What would settle it

A controlled replication would settle it: run the same camera-only and multimodal systems on the KITTI rainy split under identical conditions, and on an adversarial-patch test set, to see whether mIoU rises from 72.3% to 88.6% and misclassification falls from 89% to 12% as claimed.

Watch

Extended reading notes

Core claim

The paper's central claim is that MLLMs are a viable upgrade path for traffic safety, not just a research curiosity. It assembles comparative evidence that multimodal fusion and contextual reasoning beat isolated sensor processing: on KITTI rainy scenes, mIoU rises from 72.3% with camera-only input to 88.6% with multimodal fusion; near-miss anomaly detection reaches 76.31% accuracy versus 41.31% for rule-based baselines; adversarial misclassification falls from 89% to 12% when GPS and historical data are cross-referenced; and quantized MLLMs run at 25 FPS on edge devices, comparable to lightweight CNNs. The review also classifies MLLM applications into perception, decision-making, human interaction, and safety-critical analysis, and identifies datasets (KITTI, DRAMA, SHRP2 NDS, ML4RoadSafety, among others) that enable such systems.

Load-bearing premise

The review's comparative verdict rests on the assumption that the cited performance figures — 72.3% to 88.6% mIoU, 89% to 12% misclassification, 35% faster responses, 89% AccidentGPT precision — actually come from the cited studies and are reproducible, yet the text gives no experimental protocols, error bars, or primary-source verification for these numbers.

Editorial extensions

If this is right

  • Physical adversarial patches on road signs become far less effective: with cross-modal validation the reported misclassification rate drops from 89% to 12%, directly removing a known failure mode of camera-only detectors.
  • MLLMs can be deployed in vehicles rather than only in the cloud: quantized variants reach 25 FPS on edge hardware, comparable to lightweight convolutional baselines.
  • Interactive accident analysis becomes practical: MLLMs on the DRAMA dataset cut post-processing latency by 40% while answering natural-language queries about near-miss events.
  • Risk prediction becomes proactive: AccidentGPT-style systems report 89% precision in identifying collision-prone areas, enabling intervention before incidents occur.
  • Zero-shot capabilities reduce data dependence: near-miss detection at 76.31% accuracy works without task-specific training data, easing adoption in new traffic environments.

Reading between the lines

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

  • If the reported gains come from cross-modal validation, then synchronized multi-sensor attacks — perturbing camera and LiDAR feeds together — become the critical next test, since the cited evidence only covers single-modal attacks.
  • The numbers as presented argue for a standardized public benchmark of MLLM traffic-safety systems, with fixed attacks, weather conditions, and metrics, before any of these figures can guide deployment decisions.
  • The paper's emphasis on hallucination reduction implies that false-positive hazard warnings, not just missed detections, should be a primary safety metric, because a system that cries wolf erodes driver trust.
  • Causality-driven reasoning is the natural extension of the paper's argument, but it would require new datasets annotated with causal structure, which none of the reviewed datasets currently provide.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 6 minor

Summary. The manuscript is a narrative review claiming that multimodal large language models (MLLMs) improve traffic safety relative to traditional ADAS by integrating vision, language, and environmental signals. It presents a quantitative comparison in Table 1 (accuracy, inference speed, energy, adversarial robustness), organizes MLLM applications into perception, decision-making, human-machine interaction, and safety-critical analysis, reviews several driving datasets, and concludes with future research directions. The central claim is that MLLMs provide holistic scene understanding that outperforms conventional, isolated-sensor systems.

Significance. The topic is timely, and the proposed taxonomy in Section 3 (perception, decision-making, interaction, safety-critical analysis) is a reasonable organizing structure for a survey. However, the manuscript's main empirical evidence is not reliable: multiple quantitative claims in Table 1 and Section 3 are attributed to papers that are not MLLM systems, do not address the stated scenarios, or do not contain the reported numbers. The paper also lacks a systematic search or extraction protocol, so the comparison is not auditable. If the quantitative comparisons were corrected or removed, the remaining narrative could serve as a position piece, but in its current form the paper does not provide a trustworthy foundation for the claim that MLLMs 'enhance' traffic safety.

major comments (5)
  1. [Section 2, Table 1] The mIoU row claims an improvement from 72.3% (camera-only) to 88.6% (multimodal fusion) on KITTI under rainy conditions, but this evidence does not test the MLLM hypothesis. Source [13] is CAFuser, a condition-aware multimodal fusion network without any LLM component, and KITTI [12] has no standard rainy-condition benchmark. As written, this row neither supports the paper's central claim about MLLMs nor provides a verifiable baseline.
  2. [Section 2, Table 1] The adversarial robustness rows are unsupported by the cited sources. The 41%/79% accuracy figures cite [19], which studies physical adversarial attacks on an aerial-imagery object detector, not a camera-only ADAS or a LiDAR+thermal fusion system; the 89%/12% misclassification reduction cites [20], a paper on vision-language model robustness that does not report the claimed 'GPS + historical data cross-validation' experiment. These numbers therefore cannot be verified from the references given.
  3. [Section 3, LLaVA-ST paragraph] The claim that LLaVA-ST 'reduces localization errors by 27% on the ST-Align dataset (4.3M samples)' cites [25], which is a paper on image-gene alignment in spatial transcriptomics, not a driving dataset. Additionally, references [14], [6], and [26] all point to the same R. Zhang et al. preprint, yet the manuscript uses it for three different quantitative results (76.31% zero-shot anomaly detection, 35% improvement, and 40% latency reduction) without describing how these numbers were obtained. This is a load-bearing evidence problem for the review's empirical claims.
  4. [Section 3, AccidentGPT and hallucination-fine-tuning] The '89% precision' attributed to AccidentGPT is cited to [30] without indicating whether the figure appears in that paper or is a reinterpretation. The subsequent sentence states that Gemini-Pro-Vision 1.5 and LLaVA 'reduce hallucination errors, improving zero-shot accuracy on the SHRP2 NDS dataset' and cites [35], which is a general survey on multimodal hallucination and does not report experiments on SHRP2 NDS. These unsupported causal claims should be removed or replaced with primary sources that actually contain the reported evaluations.
  5. [Sections 2 and 3, overall methodology] The review does not disclose a systematic search strategy, inclusion/exclusion criteria, or a data-extraction protocol for the quantitative comparisons in Table 1 and the narrative in Section 3. Without such a protocol, the reader cannot distinguish a comprehensive, balanced survey from a selection of favorable results, and the specific numbers appear chosen to support the thesis rather than to represent the field. This undermines the paper's claim to be a 'comprehensive review'.
minor comments (6)
  1. [Table 1] The table heading 'Comparison of Existing Dataset for Safety Traffic' is grammatically incorrect, and the table compares ADAS and MLLM systems, not datasets; the same heading appears on Table 2.
  2. [Table 2] The NuScenes row lists '1K RGB images'; NuScenes actually contains 1,000 driving scenes with about 1.4 million camera images, so this entry is misleading.
  3. [Table 1] The 'Misclassification Rate 1K' header contains a stray '1K' that should be removed.
  4. [Section 3] The phrase 'different in-context learning (ICT) settings' should be 'in-context learning (ICL) settings'; the abbreviation appears to be a typo.
  5. [Section 3, SeeUnsafe discussion] The sentence 'we investigate other settings such as self-ensemble learning...' appears to be copied from the authors' prior work [5] and is out of place in a review; the pronoun 'we' is confusing in a survey context.
  6. [References] Several references have malformed author lists (e.g., [27], [38], [39]) and reference [14] is a duplicate of [6] and [26]; these should be unified and formatted according to the journal's style.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the review summarizes external results and performs no derivation that reduces to its own inputs.

full rationale

This is a survey, not a derivation. The paper does not fit parameters to a subset of data and then predict a closely related quantity, nor does it define a central object in terms of the conclusion. Its comparative claims are presented as summaries of external benchmarks KITTI [12], DRAMA [27], ML4RoadSafety [40], and of named systems CAFuser [13], TrafficGPT [29], AccidentGPT [30], LLaVA-ST [24], so there is no equation or construction in which an output is equal to its input. The authors do cite their own prior work [2], [3], [5], [21], [22], [43], [47], [49], [50], and one passage slides from a recent study to we investigate other settings [5], but none of these self-citations functions as an external uniqueness theorem or as the sole load-bearing support for the central claim; the central MLLMs improve traffic safety thesis is supported by a broad set of independent citations. The most serious problems in the paper are evidentiary rather than circular: some Table 1 entries labeled MLLM-Based System come from systems that are not large language models (e.g., CAFuser is a multimodal fusion network), several numbers are not traceable to primary sources, and the same reference appears under multiple numbers [6]=[14]=[26]. These are accuracy and validity concerns, not circularity, because the claimed conclusion is not made equivalent to its inputs by construction. Accordingly, no circular step is identified and the score is 0.

Assumptions & free parameters 1 free parameters · 2 assumptions · 0 invented entities

The paper introduces no new entities, forces, or quantities. It relies on existing concepts (MLLMs, datasets, ADAS components) and on the correctness of the cited performance numbers. The main epistemic burden is the accuracy of the secondary-source numbers, which is not met.

free parameters (1)
  • No free parameters
    This is a review paper with no fitted models or data fitting; the numerical values in Table 1 are borrowed from cited works (though with incomplete attribution) and are not fit here.
assumptions (2)
  • domain assumption The figures in Table 1 and the text (e.g., 72.3% mIoU, 76.31% zero-shot anomaly, 89% to 12% misclassification) accurately represent the cited works.
    The review's central comparative claims rest on these numbers, but the citations do not reliably support them. The claim about 'reduce localization errors by 27% on the ST-Align dataset' is attributed to LLaVA-ST via reference [25], which is actually a spatial-transcriptomics paper.
  • standard math The reader's aim is to understand the field of MLLMs in traffic safety, and the classification in Section 3 (perception, decision-making, HMI, safety-critical) is a useful way to organize the literature.
    This is a background assumption for any survey; it is not a mathematical or physical axiom. Standard review practice.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Multimodal Large Language Models for Enhanced Traffic Safety: A Comprehensive Review and Future Trends." pith.science (2026). https://pith.science/paper/IQJKPMZP

@misc{pith2026250416134,
  author       = {Pith},
  title        = {Pith review of: Multimodal Large Language Models for Enhanced Traffic Safety: A Comprehensive Review and Future Trends},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IQJKPMZP}},
  note         = {Machine review of arXiv:2504.16134}
}
read the original abstract

Traffic safety remains a critical global challenge, with traditional Advanced Driver-Assistance Systems (ADAS) often struggling in dynamic real-world scenarios due to fragmented sensor processing and susceptibility to adversarial conditions. This paper reviews the transformative potential of Multimodal Large Language Models (MLLMs) in addressing these limitations by integrating cross-modal data such as visual, spatial, and environmental inputs to enable holistic scene understanding. Through a comprehensive analysis of MLLM-based approaches, we highlight their capabilities in enhancing perception, decision-making, and adversarial robustness, while also examining the role of key datasets (e.g., KITTI, DRAMA, ML4RoadSafety) in advancing research. Furthermore, we outline future directions, including real-time edge deployment, causality-driven reasoning, and human-AI collaboration. By positioning MLLMs as a cornerstone for next-generation traffic safety systems, this review underscores their potential to revolutionize the field, offering scalable, context-aware solutions that proactively mitigate risks and improve overall road safety.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

57 extracted references · 33 canonical work pages

  1. [26]

    When language and vision meet road safety: leveraging multimodal large language models for video -based traffic accident analysis,

    R. Zhang, B. Wang, J. Zhang, Z. Bian, C. Feng, and K. Ozbay, “When language and vision meet road safety: leveraging multimodal large language models for video -based traffic accident analysis,” Jan. 2025

    work page 2025

  2. [13]

    CAFuser: Condition -Aware Multimodal Fusion for Robust Semantic Perception of Driving Scenes,

    T. Brödermann, C. Sakaridis, Y. Fu, and L. Van Gool, “CAFuser: Condition -Aware Multimodal Fusion for Robust Semantic Perception of Driving Scenes,” IEEE Robot Autom Lett, vol. 10, no. 4, pp. 3134–3141, Apr. 2025, doi: 10.1109/LRA.2025.3536218

    arXiv 2025

  3. [12]

    Vision meets robotics: The KITTI dataset,

    A. Geiger, P. Lenz, C. Stiller, and R. Urtasun, “Vision meets robotics: The KITTI dataset,” Int J Rob Res , vol. 32, no. 11, pp. 1231 –1237, Sep. 2013, doi: 10.1177/0278364913491297. 13

    doi:10.1177/0278364913491297 2013

  4. [19]

    Physical Adversarial Attacks on an Aerial Imagery Object Detector,

    A. Du et al., “Physical Adversarial Attacks on an Aerial Imagery Object Detector,” Aug. 2021

    work page 2021

  5. [20]

    On the Adversarial Robustness of Multi -Modal Foundation Models,

    C. Schlarmann and M. Hein, “On the Adversarial Robustness of Multi -Modal Foundation Models,” in 2023 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW) , IEEE, Oct. 2023, pp. 3679 –3687. doi: 10.1109/ICCVW60793.2023.00395

    arXiv 2023

  6. [25]

    ST-Align: A Multimodal Foundation Model for Image -Gene Alignment in Spatial Transcriptomics,

    Y. Lin et al., “ST-Align: A Multimodal Foundation Model for Image -Gene Alignment in Spatial Transcriptomics,” Nov. 2024

    work page 2024

  7. [30]

    AccidentGPT: A V2X Environmental Perception Multi -modal Large Model for Accident Analysis and Prevention,

    L. Wang et al., “AccidentGPT: A V2X Environmental Perception Multi -modal Large Model for Accident Analysis and Prevention,” in 2024 IEEE Intelligent Vehicles Symposium (IV), IEEE, Jun. 2024, pp. 472–477. doi: 10.1109/IV55156.2024.10588374. 14

    arXiv 2024

  8. [35]

    Hallucination of Multimodal Large Language Models: A Survey,

    Z. Bai et al., “Hallucination of Multimodal Large Language Models: A Survey,” Apr. 2024

    work page 2024

Show all 57 references
  1. [1]

    Road traffic Injuries

    World Health Organizatio, “Road traffic Injuries.” Accessed: Jan. 17, 2025. [Online]. Available: https://www.who.int/news-room/fact-sheets/detail/road-traffic-injuries

    2025

  2. [2]

    Leveraging Deep Learning and Multimodal Large Language Models for Near -Miss Detection Using Crowdsourced Videos,

    S. Jaradat, M. Elhenawy, H. I. Ashqar, A. Paz, and R. Nayak, “Leveraging Deep Learning and Multimodal Large Language Models for Near -Miss Detection Using Crowdsourced Videos,” IEEE Open Journal of the Computer Society, 2025

    2025

  3. [3]

    Advancing Object Detection in Transportation with Multimodal Large Language Models (MLLMs): A Comprehensive Review and Empirical Testing,

    H. I. Ashqar, A. Jaber, T. I. Alhadidi, and M. Elhenawy, “Advancing Object Detection in Transportation with Multimodal Large Language Models (MLLMs): A Comprehensive Review and Empirical Testing,” arXiv preprint arXiv:2409.18286 , 2024

    2024 arXiv

  4. [4]

    Adversarial examples: attacks and defenses in the physical world,

    H. Ren, T. Huang, and H. Yan, “Adversarial examples: attacks and defenses in the physical world,” International Journal of Machine Learning and Cybernetics , vol. 12, no. 11, pp. 3325–3336, Nov. 2021, doi: 10.1007/s13042-020-01242-z

    2021 doi

  5. [5]

    Using Multimodal Large Language Models (MLLMs) for Automated Detection of Traffic Safety-Critical Events,

    M. A. Tami, H. I. Ashqar, M. Elhenawy, S. Glaser, and A. Rakotonirainy, “Using Multimodal Large Language Models (MLLMs) for Automated Detection of Traffic Safety-Critical Events,” Vehicles, vol. 6, no. 3, pp. 1571–1590, 2024

    2024

  6. [7]

    A Comprehensive Survey of Multimodal Large Language Models: Concept, Application and Safety,

    S. Liu et al. , “A Comprehensive Survey of Multimodal Large Language Models: Concept, Application and Safety,” Oct. 18, 2024. doi: 10.21203/rs.3.rs-5270567/v1

    2024 doi

  7. [8]

    Adversarial Sensor Attack on LiDAR-based Perception in Autonomous Driving,

    Y. Cao et al., “Adversarial Sensor Attack on LiDAR-based Perception in Autonomous Driving,” in Proceedings of the 2019 ACM SIGSAC Conference on Computer and Communications Security, New York, NY, USA: ACM, Nov. 2019, pp. 2267 –2281. doi: 10.1145/3319535.3339815

    2019

  8. [9]

    MM -LLMs: Recent Advances in MultiModal Large Language Models,

    D. Zhang et al. , “MM -LLMs: Recent Advances in MultiModal Large Language Models,” in Findings of the Association for Computational Linguistics ACL 2024 , Stroudsburg, PA, USA: Association for Computational Linguistics, 2024, pp. 12401 – 12430. doi: 10.18653/v1/2024.findings-acl.738

    2024 doi

  9. [10]

    Integrating LLMs With ITS: Recent Advances, Potentials, Challenges, and Future Directions,

    D. Mahmud et al. , “Integrating LLMs With ITS: Recent Advances, Potentials, Challenges, and Future Directions,” IEEE Transactions on Intelligent Transportation Systems, pp. 1–36, 2025, doi: 10.1109/TITS.2025.3528116

    2025

  10. [11]

    A Review of Current Trends, Techniques, and Challenges in Large Language Models (LLMs),

    R. Patil and V. Gudivada, “A Review of Current Trends, Techniques, and Challenges in Large Language Models (LLMs),” Applied Sciences, vol. 14, no. 5, p. 2074, Mar. 2024, doi: 10.3390/app14052074

    2024 doi

  11. [15]

    A survey of GPT -3 family large language models including ChatGPT and GPT-4,

    K. S. Kalyan, “A survey of GPT -3 family large language models including ChatGPT and GPT-4,” Natural Language Processing Journal, vol. 6, p. 100048, Mar. 2024, doi: 10.1016/j.nlp.2023.100048

    2024

  12. [16]

    MobileVLM V2: Faster and Stronger Baseline for Vision Language Model,

    X. Chu et al. , “MobileVLM V2: Faster and Stronger Baseline for Vision Language Model,” Feb. 2024

    2024

  13. [17]

    MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications,

    A. G. Howard et al., “MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications,” Apr. 2017

    2017

  14. [18]

    Dynamic -LLaVA: Efficient Multimodal Large Language Models via Dynamic Vision -language Context Sparsification,

    Wenxuan Huang and Zijie Zhai and Yunhang Shen and Shaosheng Cao and Fei Zhao and Xiangfeng Xu and Zheyu Ye and Shaohui Lin, “Dynamic -LLaVA: Efficient Multimodal Large Language Models via Dynamic Vision -language Context Sparsification,” The Thirteenth International Conference...

    2025

  15. [21]

    Visual Reasoning and Multi-Agent Approach in Multimodal Large Language Models (MLLMs): Solving TSP and mTSP Combinatorial Challenges,

    M. Elhenawy et al., “Visual Reasoning and Multi-Agent Approach in Multimodal Large Language Models (MLLMs): Solving TSP and mTSP Combinatorial Challenges,” Mach Learn Knowl Extr , vol. 6, no. 3, pp. 1894 –1920, 2024, doi: 10.3390/make6030093

    1920 doi

  16. [22]

    Benchmarking LLMs for Real -World Applications: From Numerical Metrics to Contextual and Qualitative Evaluation,

    H. I. Ashqar, “Benchmarking LLMs for Real -World Applications: From Numerical Metrics to Contextual and Qualitative Evaluation,” Authorea Preprints, 2025

    2025

  17. [23]

    Forging Vision Foundation Models for Autonomous Driving: Challenges, Methodologies, and Opportunities,

    X. Yan et al. , “Forging Vision Foundation Models for Autonomous Driving: Challenges, Methodologies, and Opportunities,” Jan. 2024

    2024

  18. [24]

    LLaVA -ST: A Multimodal Large Language Model for Fine -Grained Spatial-Temporal Understanding,

    H. Li et al. , “LLaVA -ST: A Multimodal Large Language Model for Fine -Grained Spatial-Temporal Understanding,” Jan. 2025

    2025

  19. [27]

    DRAMA: Joint Risk Localization and Captioning in Driving,

    S. and C. C. and D. I. and C. J. H. and L. J. Malla, “DRAMA: Joint Risk Localization and Captioning in Driving,” Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), vol. 1, pp. 1043–1052, 2023

    2023

  20. [28]

    Driving with LLMs: Fusing Object -Level Vector Modality for Explainable Autonomous Driving,

    L. Chen et al. , “Driving with LLMs: Fusing Object -Level Vector Modality for Explainable Autonomous Driving,” in 2024 IEEE International Conference on Robotics and Automation (ICRA) , IEEE, May 2024, pp. 14093 –14100. doi: 10.1109/ICRA57147.2024.10611018

    2024

  21. [29]

    TrafficGPT : Viewing, processing and interacting with traffic foundation models,

    S. Zhang et al. , “TrafficGPT : Viewing, processing and interacting with traffic foundation models,” Transp Policy (Oxf) , vol. 150, pp. 95 –105, May 2024, doi: 10.1016/j.tranpol.2024.03.006

    2024 doi

  22. [31]

    Drive As You Speak: Enabling Human-Like Interaction With Large Language Models in Autonomous Vehicles,

    C. and M. Y. and C. X. and Y. W. and W. Z. Cui, “Drive As You Speak: Enabling Human-Like Interaction With Large Language Models in Autonomous Vehicles,” Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) Workshops, pp. 902–909, Jan. 2024

    2024

  23. [32]

    Probing the Robustness of Vision-Language Pretrained Models: A Multimodal Adversarial Attack Approach,

    J. Guan, T. Ding, L. Cao, L. Pan, C. Wang, and X. Zheng, “Probing the Robustness of Vision-Language Pretrained Models: A Multimodal Adversarial Attack Approach,” Aug. 2024

    2024

  24. [33]

    Gemini: A Family of Highly Capable Multimodal Models,

    Gemini Team et al., “Gemini: A Family of Highly Capable Multimodal Models,” Dec. 2023

    2023

  25. [34]

    Visual Instruction Tuning,

    H. Liu, C. Li, Q. Wu, and Y. J. Lee, “Visual Instruction Tuning,” Apr. 2023

    2023

  26. [36]

    A Cloud -Edge Collaborative Architecture for Multimodal LLMs-Based Advanced Driver Assistance Systems in IoT Networks,

    Y. Hu, D. Ye, J. Kang, M. Wu, and R. Yu, “A Cloud -Edge Collaborative Architecture for Multimodal LLMs-Based Advanced Driver Assistance Systems in IoT Networks,” IEEE Internet Things J, pp. 1–1, 2024, doi: 10.1109/JIOT.2024.3509628

    2024

  27. [37]

    SurrealDriver: Designing LLM -powered Generative Driver Agent Framework based on Human Drivers’ Driving -thinking Data,

    Y. Jin et al. , “SurrealDriver: Designing LLM -powered Generative Driver Agent Framework based on Human Drivers’ Driving -thinking Data,” in 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , IEEE, Oct. 2024, pp. 966–971. doi: 10.1109/IROS58592.2...

    2024

  28. [38]

    The Cityscapes Dataset for Semantic Urban Scene Understanding,

    M. and O. M. and R. S. and R. T. and E. M. and B. R. and F. U. and R. S. and S. B. Cordts, “The Cityscapes Dataset for Semantic Urban Scene Understanding,” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)}, vol. 6, 2016

    2016

  29. [39]

    Scalability in Perception for Autonomous Driving: Waymo Open Dataset,

    P. and K. H. and D. X. and C. A. and P. V. and T. P. and G. J. and Z. Y. and C. Y. and C. B. and V. V. and H. W. and N. J. and Z. H. and T. Sun, “Scalability in Perception for Autonomous Driving: Waymo Open Dataset,” Proceedings of the IEEE/CVF Conference on Computer Vision an...

    2020

  30. [40]

    Graph neural networks for road safety modeling: datasets and evaluations for accident analysis,

    A. , L. D. , J. H. , K. H. and Z. H. Nippani, “Graph neural networks for road safety modeling: datasets and evaluations for accident analysis,” Adv Neural Inf Process Syst, 2024

    2024

  31. [41]

    Traffic Condition Classification Model Based on Traffic‐Net,

    F. Cao, S. Chen, J. Zhong, and Y. Gao, “Traffic Condition Classification Model Based on Traffic‐Net,” Comput Intell Neurosci , vol. 2023, no. 1, Jan. 2023, doi: 10.1155/2023/7812276

    2023 doi

  32. [42]

    TrafficMOT: A Challenging Dataset for Multi -Object Tracking in Complex Traffic Scenarios,

    L. Liu et al. , “TrafficMOT: A Challenging Dataset for Multi -Object Tracking in Complex Traffic Scenarios,” in Proceedings of the 32nd ACM International Conference on Multimedia , New York, NY, USA: ACM, Oct. 2024, pp. 1265 –1273. doi: 10.1145/3664647.3681153

    2024

  33. [43]

    Leveraging Multimodal Large Language Models (MLLMs) for Enhanced Object Detection and Scene Understanding in Thermal Images for Autonomous Driving Systems,

    H. I. Ashqar, T. I. Alhadidi, M. Elhenawy, and N. O. Khanfar, “Leveraging Multimodal Large Language Models (MLLMs) for Enhanced Object Detection and Scene Understanding in Thermal Images for Autonomous Driving Systems,” Automation, vol. 5, no. 4, pp. 508–526, 2024

    2024

  34. [44]

    TAD: A Large -Scale Benchmark for Traffic Accidents Detection From Video Surveillance,

    Y. Xu et al., “TAD: A Large -Scale Benchmark for Traffic Accidents Detection From Video Surveillance,” IEEE Access , vol. 13, pp. 2018 –2033, 2025, doi: 10.1109/ACCESS.2024.3522384

    2018

  35. [45]

    Description of the SHRP 2 naturalistic database and the crash, near -crash, and baseline data sets,

    J. M. , P. M. A. and M. J. A. Hankey, “Description of the SHRP 2 naturalistic database and the crash, near -crash, and baseline data sets,” Virginia Tech Transportation Institute, 2016

    2016

  36. [46]

    ScVLM: Enhancing Vision-Language Model for Safety-Critical Event Understanding,

    L. Shi, B. Jiang, T. Zeng, and F. Guo, “ScVLM: Enhancing Vision-Language Model for Safety-Critical Event Understanding,” Oct. 2024

    2024

  37. [47]

    Vision-Language Models for Autonomous Driving: CLIP -Based Dynamic Scene Understanding,

    M. Elhenawy, H. I. Ashqar, A. Rakotonirainy, T. I. Alhadidi, A. Jaber, and M. A. Tami, “Vision-Language Models for Autonomous Driving: CLIP -Based Dynamic Scene Understanding,” arXiv preprint arXiv:2501.05566, 2025. 15

    2025 arXiv

  38. [48]

    A Causality -Aware Paradigm for Evaluating Creativity of Multimodal Large Language Models,

    Z. Huang, S. Zhong, P. Zhou, S. Gao, M. Zitnik, and L. Lin, “A Causality -Aware Paradigm for Evaluating Creativity of Multimodal Large Language Models,” IEEE Trans Pattern Anal Mach Intell, pp. 1–17, 2025, doi: 10.1109/TPAMI.2025.3539433

    2025

  39. [49]

    Large Language Models (LLMs) as Traffic Control Systems at Urban Intersections: A New Paradigm,

    S. Masri, H. I. Ashqar, and M. Elhenawy, “Large Language Models (LLMs) as Traffic Control Systems at Urban Intersections: A New Paradigm,” Vehicles, vol. 7, no. 1, p. 11, 2025

    2025

  40. [50]

    Leveraging Large Language Models (LLMs) for Traffic Management at Urban Intersections: The Case of Mixed Traffic Scenarios,

    S. Masri, H. I. Ashqar, and M. Elhenawy, “Leveraging Large Language Models (LLMs) for Traffic Management at Urban Intersections: The Case of Mixed Traffic Scenarios,” arXiv preprint arXiv:2408.00948, 2024

    2024 arXiv

  41. [51]

    On Fairness of Unified Multimodal Large Language Model for Image Generation,

    M. Liu, H. Chen, J. Wang, L. Wang, B. R. Ramakrishnan, and W. Zhang, “On Fairness of Unified Multimodal Large Language Model for Image Generation,” Feb. 2025

    2025

  42. [52]

    MM-SafetyBench: A Benchmark for Safety Evaluation of Multimodal Large Language Models,

    X. Liu, Y. Zhu, J. Gu, Y. Lan, C. Yang, and Y. Qiao, “MM-SafetyBench: A Benchmark for Safety Evaluation of Multimodal Large Language Models,” 2025, pp. 386–403. doi: 10.1007/978-3-031-72992-8_22

    2025 doi

  43. [53]

    Shielded Representations: Protecting Sensitive Attributes Through Iterative Gradient -Based Projection,

    S. Iskander, K. Radinsky, and Y. Belinkov, “Shielded Representations: Protecting Sensitive Attributes Through Iterative Gradient -Based Projection,” in Findings of the Association for Computational Linguistics: ACL 2023 , Stroudsburg, PA, USA: Association for Computational Lin...

    2023 doi

  44. [54]

    MLLM -Protector: Ensuring MLLM’s Safety without Hurting Performance,

    R. Pi et al. , “MLLM -Protector: Ensuring MLLM’s Safety without Hurting Performance,” in Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing , Stroudsburg, PA, USA: Association for Computational Linguistics, 2024, pp. 16012–16027. doi: 10.186...

    2024 doi

  45. [55]

    AdaShield : Safeguarding Multimodal Large Language Models from Structure-Based Attack via Adaptive Shield Prompting,

    Y. Wang, X. Liu, Y. Li, M. Chen, and C. Xiao, “AdaShield : Safeguarding Multimodal Large Language Models from Structure-Based Attack via Adaptive Shield Prompting,” 2025, pp. 77–94. doi: 10.1007/978-3-031-72661-3_5

    2025 doi

  46. [56]

    Towards More Robust Retrieval- Augmented Generation: Evaluating RAG Under Adversarial Poisoning Attacks,

    J. Su, J. P. Zhou, Z. Zhang, P. Nakov, and C. Cardie, “Towards More Robust Retrieval- Augmented Generation: Evaluating RAG Under Adversarial Poisoning Attacks,” Dec. 2024

    2024

  47. [57]

    ACEA Position Paper Artificial Intelligence in the automobile industry,

    ACEA, “ACEA Position Paper Artificial Intelligence in the automobile industry,” https://www.acea.auto/files/ACEA_Position_Paper- Artificial_Intelligence_in_the_automotive_industry.pdf

  48. [58]

    Cross-Domain Few- Shot In -Context Learning For Enhancing Traffic Sign Recognition,

    Y. Gan, G. Li, R. Togo, K. Maeda, T. Ogawa, and M. Haseyama, “Cross-Domain Few- Shot In -Context Learning For Enhancing Traffic Sign Recognition,” in 2024 IEEE International Conference on Image Processing (ICIP) , IEEE, Oct. 2024, pp. 2564 –

    2024

  49. [2570]

    doi: 10.1109/ICIP51287.2024.10647129

    2024

Pith tools

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