Pith. sign in

REVIEW 4 major objections 5 minor 3 cited by

AI2-Active Safety: AI-enabled Interaction-aware Active Safety Analysis with Vehicle Dynamics

T0 review · 4 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read A safety framework that fuses high-fidelity bicycle dynamics, hypergraph-based group interaction modeling, and probabilistic trajectory prediction, producing a high-fidelity time-to-collision (HF-TTC) that the paper reports alerts earlier…

desk verdict Solid trajectory-forecasting core, but the headline HF-TTC is a proximity measure, not a time-to-collision, so the early-warning claims rest on a flawed metric. read the letter →

arxiv 2505.00322 v1 pith:HTABHQDP submitted 2025-05-01 cs.RO cs.AI

Keshu Wu , Zihao Li , Sixu Li , Xinyue Ye , Dominique Lord , Yang Zhou This is my paper · ORCID
classification cs.ROcs.AI
keywords activesafetyanalysissurrogatemeasurestime-to-collisionhypergraphneuralnetworktrajectorypredictionvehicledynamicsbicyclemodeluncertaintyquantification
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

The paper sets out to establish that surrogate safety analysis is improved by replacing constant-velocity extrapolation with three fused ingredients: a bicycle model of the host vehicle that includes road gradient, a hypergraph neural network that reasons about multi-vehicle groups rather than pairwise relations, and probabilistic trajectory forecasts that carry mode probabilities. From these, it derives a stochastic high-fidelity time-to-collision (HF-TTC) distribution, and reports that on the NGSIM and HighD highway datasets it issues warnings earlier than traditional constant-velocity TTC while keeping predicted collision probabilities low. If the framework works as claimed, active safety systems would have a physically grounded, uncertainty-aware risk signal that reflects group maneuvers such as coordinated lane changes and platoons, instead of a single deterministic extrapolation. The paper also reports fifty-frame trajectory prediction RMSE of 2.75 m on NGSIM and 0.41 m on HighD for its hypergraph transformer.

What carries the argument

The load-bearing object is the stochastic HF-TTC distribution. The host vehicle's future is generated by an augmented bicycle model with state $(p_x, p_y, \psi, v)$ and dynamics $\dot p_x = v\cos\psi$, $\dot p_y = v\sin\psi$, $\dot\psi = v\tan\delta/L$, $\dot v = a - g\sin\alpha$, integrated by fourth-order Runge–Kutta under several control-input hypotheses. In parallel, a hypergraph transformer infers group structure from cosine similarities of historical trajectory embeddings and outputs $M$ ambient trajectory modes with probabilities $p_m$. For each host–ambient pair and mode, a collision is identified at the first time $|\Delta x| \le r_x$ and $|\Delta y| \le r_y$, and the inverse TTC values are aggregated into a probability mass function and CDF. The work of this machinery is to turn raw trajectory data into a probability-weighted collision-time distribution that can be thresholded for early warnings.

What would settle it

Simulate a host and an ambient vehicle driving side by side in adjacent lanes with constant nonzero spacing and identical speed, with $|\Delta x| \le r_x$ and $|\Delta y| \le r_y$ already satisfied at $t_0$. Equation (38) returns HF-TTC $= t_0$ even though the distance is not shrinking; this single case shows the metric responds to proximity, not to time-to-contact. A metric that instead requires $\frac{d}{dt}|\Delta x| < 0$ or $\frac{d}{dt}|\Delta y| < 0$ before counting a collision would not fire in this scenario.

Watch

Extended reading notes

Core claim

The central claim is that the paper's HF-TTC metric—computed by running deterministic RK4 host trajectories against each probabilistic ambient trajectory mode and taking the first time both longitudinal and lateral separations fall below thresholds—is a more faithful and earlier warning signal than traditional constant-velocity TTC. The paper reports that HF-TTC consistently issues alerts well before constant-velocity TTC, that its cumulative distributions concentrate below 0.5% collision probability by five seconds while traditional TTC rises to about 3%, and that these low probabilities align with the rarity of real highway crashes. It further claims that the hypergraph transformer, by representing group interactions as hyperedges, reduces fifty-frame RMSE to 2.75 m on NGSIM and 0.41 m on HighD, and that ablations removing hypergraph reasoning, probabilistic decoding, or bicycle dynamics degrade long-horizon accuracy. The paper presents this as the first surrogate safety framework integrating high-fidelity dynamics, hypergraph interaction modeling, and stochastic trajectory prediction.

Load-bearing premise

The load-bearing premise is that a collision has occurred when both absolute separations fall below fixed thresholds, so the metric treats close proximity as collision risk even when the two vehicles are not approaching each other; if the intended meaning of HF-TTC is time to contact with an approaching threat, this definition needs a closing-velocity condition.

Editorial extensions

If this is right

  • Alert thresholds can be set from the HF-TTC CDF instead of a single deterministic number, allowing operators to balance false alarms against missed detections.
  • Group maneuvers such as coordinated lane changes and platoon decelerations enter the risk metric directly through hyperedges, so warnings can fire when a group, not just a single pair, is moving into conflict.
  • Long-horizon warnings become more trustworthy because the bicycle model and RK4 integration keep host trajectories physically consistent, which the ablation suggests matters most at 50 frames.
  • Traditional constant-velocity TTC's inflated collision probabilities (about 3% by five seconds in the paper's comparison) would be replaced by lower, empirically more plausible probabilities, reducing desensitization to real threats.
  • The same pipeline can be re-run under different host behavior models, letting the safety system compare how much risk is reduced by steering or braking choices.

Reading between the lines

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

  • My inference: the collision definition in Eq. (37)–(38) triggers on absolute separation thresholds, so a vehicle already beside the host in an adjacent lane at constant spacing would receive a near-zero HF-TTC even with zero closing speed; the metric therefore measures proximity rather than time-to-contact unless a closing-velocity condition is added.
  • My inference: the reported early-warning advantage is tied to the quality of the hypergraph predictor; if the predictor's multi-modal distribution is miscalibrated on unseen road geometries, the HF-TTC CDF will inherit that miscalibration, so the framework's safety guarantee is only as strong as its uncertainty calibration.
  • My inference: one natural extension is to use HF-ITTC as an objective in motion planning, selecting host control inputs that minimize expected inverse TTC; the paper's scenario analysis shows the host's RK4 plan changing as risk shifts between vehicles, which is the precursor to such closed-loop use.
  • My inference: because the paper only evaluates on highways with relatively structured lane changes, a stress test in merging or intersection scenarios with crossing paths would reveal whether the longitudinal and lateral separation thresholds capture the conflicts that matter there.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes an "AI2-Active Safety" framework that combines three components: an augmented bicycle model with road-gradient effects for the host vehicle, a hypergraph transformer that predicts multi-modal probabilistic trajectories of ambient vehicles, and a stochastic high-fidelity time-to-collision (HF-TTC) metric computed from the predicted trajectory distribution and RK4-integrated host trajectories. The authors evaluate the prediction component on the NGSIM and HighD datasets, compare three host-behavior models, perform ablations, and claim that HF-TTC consistently issues earlier and more realistic warnings than traditional constant-velocity TTC.

Significance. If the central safety metric were correctly defined, the integration of group-wise interaction modeling, vehicle dynamics, and uncertainty quantification into a surrogate safety measure would be a meaningful contribution to active-safety analysis. The trajectory-prediction component is independently grounded in public benchmark comparisons and reports substantial RMSE reductions (e.g., 2.75 m vs. 2.97 m at 50 frames on NGSIM compared with RHINO, and 0.41 m vs. 0.89 m on HighD), and the ablation study supports the contribution of each design choice. However, as defined, HF-TTC is a proximity/encroachment time rather than a time-to-collision, so the paper's headline early-warning claims and the quantitative comparisons against constant-velocity TTC are not currently supported. The conclusion candidly acknowledges several implementation limitations, but it does not address the definitional issue in the safety metric.

major comments (4)
  1. [§3.3, Eqs. (37)–(38)] The HF-TTC definition uses absolute separations |Δx|≤rx and |Δy|≤ry with no condition that the separation is decreasing. Consequently, a vehicle already in an adjacent lane within the lateral threshold, or a follower at constant spacing below rx, has HF-TTC=0 at t0 even when the gap is not closing, and parallel traffic with a small constant separation yields small or zero HF-TTC indefinitely. The metric therefore measures persistent proximity, not time-to-contact. Since CV-TTC only produces finite values when the range is closing, the comparisons in §4.4 and Figures 8–9 are not apples-to-apples, and the conclusion that HF-TTC "consistently issues alerts well before traditional constant-velocity TTC" inherits this artifact. I recommend adding an explicit closing-velocity condition (e.g., d|Δx|/dt < 0 and d|Δy|/dt < 0, or a negative range-rate) before recording a crossing, and re-running the scenario analyses.
  2. [§3.3, Eq. (39)] The inverse TTC is defined as HF-ITTC(m) = 1/HF-ITTC(m)(t0), which is self-referential and dimensionally inconsistent. It should presumably be 1/HF-TTC(m)(t0). This typo propagates into the PMF/CDF definitions in Eqs. (40)–(41) and into the inverse-TTC plots in Figures 8–9, so it must be corrected and all downstream formulas rechecked.
  3. [§3.3, Eq. (38)] The minimization is over t≥t0 with no upper bound, but the predicted trajectories are only defined on [t0, t0+Tp]. For a mode that never breaches both thresholds within the prediction horizon, HF-TTC is undefined; the manuscript does not specify a convention (e.g., +∞, horizon time, or censoring). Without this convention, the PMF and CDF aggregation in Eqs. (40)–(41) is not well-defined. Please clarify the convention and describe how the implementation handles non-crossing modes.
  4. [§3.2.4, Eqs. (34) and (40)–(41)] The softmax probabilities pm are treated as calibrated probabilities of future modes and used directly as the probability masses of the HF-ITTC distribution. No calibration analysis (e.g., reliability diagrams, expected calibration error) is provided. If the softmax outputs are not calibrated, the claimed "faithful" uncertainty quantification and the quantitative probability claims (e.g., "below 0.5%" in §4.4) are not justified. Please add calibration evidence or explicitly state that pm are uncalibrated model confidences and discuss the implications for setting alert thresholds.
minor comments (5)
  1. [Abstract, §1] The claim of being "the first SSM framework that seamlessly integrates" the three components is a strong novelty assertion; unless a systematic literature search supports uniqueness, I recommend tempering it to "a framework that integrates...".
  2. [Table 1, Figure 6] The reported prediction-error improvements over baselines lack error bars, confidence intervals, or significance tests across multiple seeds; reporting means±standard deviations over at least three seeds would strengthen the evaluation.
  3. [§4.4, final paragraph] The statement that CV-TTC reaching "3% by the five-second mark" is a collision probability that "contradicts the well-known rarity of highway crashes" conflates a threshold-crossing frequency with a crash probability; TTC values below a threshold are near-miss indicators, not probabilities of collision.
  4. [§3.1.1] Typo: "Cartersian" should be "Cartesian".
  5. [§3.2.4, Eq. (33)] The phrase "maps Inputi" is undefined; it should refer to the decoder input hdec_i defined in Eq. (32).

Circularity Check

1 steps flagged · score 1.0 of 10

No significant circularity in the central derivation: trajectory prediction is externally benchmarked, and HF-TTC is a forward calculation. One self-referential formula typo appears in the HF-ITTC definition.

  1. self definitional [Section 3.3, Eq. (39)]
    "The corresponding inverse TTC (HF-ITTC) is HF-ITTC(m)i j = 1 / HF-ITTC(m)i j (t0). (39)"

    As typeset, HF-ITTC is defined as the reciprocal of itself. The inverse-TTC distribution used in Section 4 cannot be computed from this equation because the intended quantity (presumably 1/HF-TTC) is absent. This is a self-referential placeholder rather than a derivation, but it is a typographical/definitional defect, not a load-bearing reduction of the paper's main results.

full rationale

The paper's central prediction claim is not circular: the Hypergraph Transformer is trained on a 70% split and evaluated on a held-out 30% split of the public NGSIM and HighD datasets, with RMSE/ADE/FDE/MAE reported against several external baselines and a constant-speed model. The ablation study compares HGT to HGT-GNN, HGT-Deterministic, and HGT-Kinematic, none of which are fitted to the test set. The HF-TTC computation is a forward calculation from predicted trajectory modes and RK4 host trajectories; it is not a fitted parameter renamed as a prediction. The paper does not rely on a self-citation chain to force its architecture: the hypergraph transformer is built on an externally published MART baseline [27], and the authors' own prior hypergraph work (RHINO) is used only as a baseline to beat. The main caveats are substantive validity issues rather than circularity: the HF-TTC definition of Eq. (38) is a threshold-encroachment time without a closing-velocity condition, and the softmax probabilities pm are treated as collision probabilities without calibration. These affect the correctness of the safety-metric claim but do not make the derivation equivalent to its inputs. The only literal self-reference found in the derivation chain is Eq. (39), where HF-ITTC is written as the reciprocal of itself; because the intended reciprocal-of-HF-TTC is obvious and the trajectory-prediction results are independently benchmarked, this yields a minimal circularity score.

Assumptions & free parameters 6 free parameters · 5 assumptions · 2 invented entities

The central contribution rests on several hand-set quantities (τ, rx, ry, M, K) and uncalibrated softmax probabilities. The hypergraph transformer and the HF-TTC metric are new constructs introduced by this paper, but neither is released in a reproducible form, so the reader cannot independently verify them.

free parameters (6)
  • Hypergraph affinity threshold τ = not reported
    Eq. (22) binarizes the cosine-similarity matrix; its value changes hypergraph density and all downstream risk estimates, yet no value or sensitivity analysis is given.
  • Longitudinal safety threshold rx = not reported
    Used in Eq. (37) to declare collision; HF-TTC values are directly set by this threshold.
  • Lateral safety threshold ry = not reported
    Used in Eq. (37); same role as rx.
  • Prediction mode count M = not reported (figures show 3)
    Number of decoder heads in Eq. (33); directly controls the shape of the HF-TTC distribution.
  • Control-sequence count K = 3 (implicitly)
    The paper lists three behavior hypotheses: Last-Step Constant, Average Constant, Self-Prediction.
  • Time step Δt and horizon Tp = not reported
    Tp corresponds to 50 forecast frames at dataset sample rates; Δt is not explicitly given.
assumptions (5)
  • domain assumption Bicycle model with rigid-body no-slip kinematics and constant wheelbase L (Eq. 2)
    Assumes planar kinematics for the host vehicle; ignores lateral tire forces, load transfer, and roll.
  • domain assumption Road gradient enters only the longitudinal acceleration as g sin(α) (Eq. 3)
    Neglects slope effects on steering, tire grip, and vertical road geometry; the '3D road surface' claim rests on this single term.
  • ad hoc to paper Hyperedge membership inferred from cosine similarity of trajectory embeddings above threshold τ (Eqs. 21-22)
    No evidence that this heuristic recovers true interaction groups; it grounds all group-wise predictions.
  • ad hoc to paper Softmax probabilities pm are treated as calibrated likelihoods of future modes (Eq. 34)
    The paper interprets resulting probability masses as collision-risk probabilities, but softmax logits are not calibrated.
  • domain assumption Collision occurs when |Δx|≤rx and |Δy|≤ry simultaneously (Eqs. 37-38)
    TTC is a time-to-threshold-entry, not a time-to-contact; no closing-speed condition.
invented entities (2)
  • Hypergraph Transformer (HGT)
    purpose: Learns group-wise interactions and predicts multimodal trajectory distributions for all ambient vehicles.
    A new model architecture in this paper, adapted from MART [27]; no code or weights are released, so its behavior outside this paper cannot be independently reproduced from the preprint alone.
  • High-fidelity TTC (HF-TTC) metric
    purpose: Surrogate safety measure that combines continuous host dynamics with probabilistic ambient trajectories.
    Introduced here as an extension of earlier group safety metrics; it is not validated against conflict or crash data, and its collision definition has the proximity issue described in the weakest assumption.

how reviews work

0 comments
Cite this review

Pith. "Pith review of AI2-Active Safety: AI-enabled Interaction-aware Active Safety Analysis with Vehicle Dynamics." pith.science (2026). https://pith.science/paper/HTABHQDP

@misc{pith2026250500322,
  author       = {Pith},
  title        = {Pith review of: AI2-Active Safety: AI-enabled Interaction-aware Active Safety Analysis with Vehicle Dynamics},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HTABHQDP}},
  note         = {Machine review of arXiv:2505.00322}
}
read the original abstract

This paper introduces an AI-enabled, interaction-aware active safety analysis framework that accounts for groupwise vehicle interactions. Specifically, the framework employs a bicycle model-augmented with road gradient considerations-to accurately capture vehicle dynamics. In parallel, a hypergraph-based AI model is developed to predict probabilistic trajectories of ambient traffic. By integrating these two components, the framework derives vehicle intra-spacing over a 3D road surface as the solution of a stochastic ordinary differential equation, yielding high-fidelity surrogate safety measures such as time-to-collision (TTC). To demonstrate its effectiveness, the framework is analyzed using stochastic numerical methods comprising 4th-order Runge-Kutta integration and AI inference, generating probability-weighted high-fidelity TTC (HF-TTC) distributions that reflect complex multi-agent maneuvers and behavioral uncertainties. Evaluated with HF-TTC against traditional constant-velocity TTC and non-interaction-aware approaches on highway datasets, the proposed framework offers a systematic methodology for active safety analysis with enhanced potential for improving safety perception in complex traffic environments.

Figures

Figures reproduced from arXiv: 2505.00322 by the authors.

Figure 1
Figure 1. Framework Architecture [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Hypergraph Transformer Network. 3.2.3. Hypergraph Representation A hypergraph is a natural generalization of a conventional graph, in which a hyperedge can connect more than two vertices. In our framework, vehicles are modeled as vertices and their group-wise interactions—as inferred from historical trajectory similarities—are captured as hyperedges. Formally, a hypergraph is de￾fined as H = {V, E, A}, (13) where V … view at source ↗
Figure 3
Figure 3. Graph and hypergraph. interactions among vehicles, and A represents the adjacency matrix that quantifies these interactions based on shared hyperedges. One common way to represent a hypergraph mathematically is by means of its incidence matrix H ∈ {0, 1} |V|×|E|. The entry H(v, e) =    1, if v ∈ e, 0, otherwise, (14) indicates whether vertex v belongs to hyperedge e. For two distinct vertices vi and vj , the … view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Hypergraph Relational Transformer. h (l) j through a series of transformer layers [36]. Each layer leverages the inferred topology G to integrate both individual and group-level information. In each transformer layer l, the update mechanism begins with an augmented sel…
Figure 5
Figure 5. Figure 5: Stochastic High-fidelity Time-to-Collision. [PITH_FULL_IMAGE:figures/full_fig_p012_5.png]
Figure 6
Figure 6. Figure 6: Comparison of positional prediction error distributions of HGT. [PITH_FULL_IMAGE:figures/full_fig_p015_6.png]
Figure 7
Figure 7. Figure 7: Scenario Analysis: host vehicle and surrounding vehicle trajectories (top row), interaction and incidence matrices (top [PITH_FULL_IMAGE:figures/full_fig_p016_7.png]
Figure 8
Figure 8. Figure 8: Comparison of host behavior models in a dense tra [PITH_FULL_IMAGE:figures/full_fig_p018_8.png]
Figure 9
Figure 9. Figure 9: Cumulative distribution of TTC across the di [PITH_FULL_IMAGE:figures/full_fig_p019_9.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 3 Pith papers

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

  1. Forecasting the Emergence and Evolution of Crash Hotspots: A Unified Deep Learning Framework for Proactive Traffic Safety

    cs.LG 2026-07 conditional novelty 5.0 of 10

    A CNN–Transformer with exposure anchoring, Hawkes self-excitation, and mixture-of-experts forecasts weekly crash hotspots and life cycles better than five matched baselines across six Wisconsin counties.

  2. Simulating the Unseen: Crash Prediction Must Learn from What Did Not Happen

    cs.LG 2025-05 conditional novelty 4.0 of 10

    Crash prediction should learn from near-miss events and synthetic counterfactual scenarios, not just recorded crashes.

  3. Automated Vehicles Should be Connected with Natural Language

    cs.MA 2025-06 conditional novelty 3.0 of 10

    A vision paper recommending natural language as the universal communication medium for connected and automated vehicles.

Reference graph

Works this paper leans on

45 extracted references · 39 canonical work pages · cited by 3 Pith papers

  1. [1]

    Estimating the expected number of crashes with tra ffic conflicts and the lomax distribution–a theoretical and numerical exploration,

    A. P. Tarko, “Estimating the expected number of crashes with tra ffic conflicts and the lomax distribution–a theoretical and numerical exploration,” Accident Analysis & Prevention, vol. 113, pp. 63–73, 2018

    work page 2018

  2. [2]

    Surrogate measures of safety,

    A. P. Tarko, “Surrogate measures of safety,” in Safe mobility: challenges, methodology and solutions, pp. 383–405, Emerald Publishing Limited, 2018

    work page 2018

  3. [3]

    Fuzzy surrogate safety metrics for real-time assessment of rear-end collision risk. a study based on empirical observations,

    K. Mattas, M. Makridis, G. Botzoris, A. Kriston, F. Minarini, B. Papadopoulos, F. Re, G. Rognelund, and B. Ciu ffo, “Fuzzy surrogate safety metrics for real-time assessment of rear-end collision risk. a study based on empirical observations,”Accident Analysis & Prevention, vol. 148, p. 105794, 2020

    work page 2020

  4. [4]

    A review of surrogate safety measures and their applications in connected and automated vehicles safety modeling,

    C. Wang, Y . Xie, H. Huang, and P. Liu, “A review of surrogate safety measures and their applications in connected and automated vehicles safety modeling,” Accident Analysis & Prevention, vol. 157, p. 106157, 2021

    2021

  5. [5]

    Studying the safety impact of autonomous vehicles using simulation- based surrogate safety measures,

    M. M. Morando, Q. Tian, L. T. Truong, and H. L. Vu, “Studying the safety impact of autonomous vehicles using simulation- based surrogate safety measures,” Journal of advanced transportation, vol. 2018, no. 1, p. 6135183, 2018

    work page 2018

  6. [6]

    Surrogate safety assessment model and validation,

    D. Gettman, L. Pu, T. Sayed, S. G. Shelby, et al., “Surrogate safety assessment model and validation,” tech. rep., Turner- Fairbank Highway Research Center, 2008

    work page 2008

  7. [7]

    Near miss determination through use of a scale of danger,

    J. C. Hayward, “Near miss determination through use of a scale of danger,” 1972

    1972

  8. [8]

    Derivation and validation of new simulation-based surrogate safety measure,

    K. Ozbay, H. Yang, B. Bartin, and S. Mudigonda, “Derivation and validation of new simulation-based surrogate safety measure,” Transportation research record, vol. 2083, no. 1, pp. 105–113, 2008

    work page 2008

Show all 45 references
  1. [9]

    Analysis of traffic conflicts and collisions,

    B. L. Allen, B. T. Shin, and P. J. Cooper, “Analysis of traffic conflicts and collisions,” tech. rep., 1978

    1978

  2. [10]

    Learning the representation of surrogate safety measures to identify tra ffic conflict,

    J. Lu, O. Grembek, and M. Hansen, “Learning the representation of surrogate safety measures to identify tra ffic conflict,” Accident Analysis & Prevention, vol. 174, p. 106755, 2022

    2022

  3. [11]

    Safety evaluation for connected and autonomous vehicles’ exclusive lanes considering penetrate ratios and impact of trucks using surrogate safety measures,

    J. Zhang, K. Wu, M. Cheng, M. Yang, Y . Cheng, and S. Li, “Safety evaluation for connected and autonomous vehicles’ exclusive lanes considering penetrate ratios and impact of trucks using surrogate safety measures,” Journal of advanced transportation, vol. 2020, pp. 1–16, 2020

    2020

  4. [12]

    An automated surrogate safety analysis at protected highway ramps using cross-sectional and before–after video data,

    P. St-Aubin, L. Miranda-Moreno, and N. Saunier, “An automated surrogate safety analysis at protected highway ramps using cross-sectional and before–after video data,” Transportation Research Part C: Emerging Technologies, vol. 36, pp. 284–295, 2013

    2013

  5. [13]

    Explicitly incorporating surrogate safety measures into connected and automated vehicle longitudinal control objectives for enhancing platoon safety,

    Y . Dai, C. Wang, and Y . Xie, “Explicitly incorporating surrogate safety measures into connected and automated vehicle longitudinal control objectives for enhancing platoon safety,”Accident Analysis & Prevention, vol. 183, p. 106975, 2023

    2023

  6. [14]

    Longitudinal safety evaluation of connected vehicles’ platooning on expressways,

    M. S. Rahman and M. Abdel-Aty, “Longitudinal safety evaluation of connected vehicles’ platooning on expressways,” Acci- dent Analysis & Prevention, vol. 117, pp. 381–391, 2018

    2018

  7. [15]

    Vehicle trajectory data extraction from the horizontal curves of mountainous roads,

    V . B. K. Anna, S. P. Venthuruthiyil, and M. Chunchu, “Vehicle trajectory data extraction from the horizontal curves of mountainous roads,” Transportation Letters, 2023

    2023

  8. [16]

    Adaptive cruise control under threat: A stochastic active safety analysis of sensing attacks in mixed traffic,

    Z. Li, Y . Zhou, J. Jiang, Y . Zhang, and M. M. Kulkarni, “Adaptive cruise control under threat: A stochastic active safety analysis of sensing attacks in mixed traffic,” Accident Analysis & Prevention, vol. 209, p. 107813, 2025

    2025

  9. [17]

    Longitudinal tra ffic conflict analysis of autonomous and traditional vehicle platoons in field tests via surrogate safety measures,

    T. Das, M. S. Samandar, and N. Rouphail, “Longitudinal tra ffic conflict analysis of autonomous and traditional vehicle platoons in field tests via surrogate safety measures,” Accident Analysis & Prevention, vol. 177, p. 106822, 2022

    2022

  10. [18]

    A joint-probability approach to crash prediction models,

    X. Pei, S. Wong, and N.-N. Sze, “A joint-probability approach to crash prediction models,” Accident Analysis & Prevention, vol. 43, no. 3, pp. 1160–1166, 2011

    2011

  11. [19]

    Beyond 1d and oversimplified kinematics: A generic analytical framework for surrogate safety measures,

    S. Li, M. Anis, D. Lord, H. Zhang, Y . Zhou, and X. Ye, “Beyond 1d and oversimplified kinematics: A generic analytical framework for surrogate safety measures,” Accident Analysis & Prevention, vol. 204, p. 107649, 2024

    2024

  12. [20]

    What do surrogate safety metrics measure? understanding driving safety as a continuum,

    A. Y . Bin-Nun, C. Lizarazo, A. Panasci, S. Madden, and R. J. D. Tebbens, “What do surrogate safety metrics measure? understanding driving safety as a continuum,” Accident Analysis & Prevention, vol. 195, p. 107245, 2024

    2024

  13. [21]

    A relative approach to the validation of surrogate measures of safety,

    C. Johnsson, A. Laureshyn, and C. D´agostino, “A relative approach to the validation of surrogate measures of safety,”Accident Analysis & Prevention, vol. 161, p. 106350, 2021

    2021

  14. [22]

    Probabilistic analysis of dynamic scenes and collision risks assessment to improve driving safety,

    C. Laugier, I. E. Paromtchik, M. Perrollaz, M. Yong, J.-D. Yoder, C. Tay, K. Mekhnacha, and A. N `egre, “Probabilistic analysis of dynamic scenes and collision risks assessment to improve driving safety,”IEEE Intelligent Transportation Systems Magazine, vol. 3, no. 4, pp. 4–19, 2011

    2011

  15. [23]

    Multimodal driver state modeling through unsupervised learning,

    A. Tavakoli and A. Heydarian, “Multimodal driver state modeling through unsupervised learning,” Accident Analysis & Prevention, vol. 170, p. 106640, 2022

    2022

  16. [24]

    A novel multimodal vehicle path prediction method based on temporal convolutional networks,

    M. N. Azadani and A. Boukerche, “A novel multimodal vehicle path prediction method based on temporal convolutional networks,” IEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 12, pp. 25384–25395, 2022

    2022

  17. [25]

    Hypergraph-based motion generation with multi-modal interaction relational reasoning,

    K. Wu, Y . Zhou, H. Shi, D. Lord, B. Ran, and X. Ye, “Hypergraph-based motion generation with multi-modal interaction relational reasoning,” arXiv preprint arXiv:2409.11676, 2024

    2024 arXiv

  18. [26]

    Disturbances and safety analysis of linear adaptive cruise control for cut-in scenarios: A theoretical framework,

    Z. Li, Y . Zhou, D. Chen, and Y . Zhang, “Disturbances and safety analysis of linear adaptive cruise control for cut-in scenarios: A theoretical framework,” Transportation Research Part C: Emerging Technologies, vol. 168, p. 104576, 2024. 22

    2024

  19. [27]

    Mart: Multiscale relational transformer networks for multi-agent trajectory prediction,

    S. Lee, J. Lee, Y . Yu, T. Kim, and K. Lee, “Mart: Multiscale relational transformer networks for multi-agent trajectory prediction,” in European Conference on Computer Vision, pp. 89–107, Springer, 2024

    2024

  20. [28]

    Interaction-aware motion planning for autonomous vehicles with multi-modal obstacle uncertainty predictions,

    J. Zhou, B. Olofsson, and E. Frisk, “Interaction-aware motion planning for autonomous vehicles with multi-modal obstacle uncertainty predictions,” IEEE Transactions on Intelligent Vehicles, vol. 9, no. 1, pp. 1305–1319, 2023

    2023

  21. [29]

    Graph-based interaction-aware multimodal 2d vehicle trajectory prediction using diffusion graph convolutional networks,

    K. Wu, Y . Zhou, H. Shi, X. Li, and B. Ran, “Graph-based interaction-aware multimodal 2d vehicle trajectory prediction using diffusion graph convolutional networks,”IEEE Transactions on Intelligent Vehicles, vol. 9, no. 2, pp. 3630–3643, 2023

    2023

  22. [30]

    Multi-modal integrated prediction and decision-making with adaptive interaction modality explorations,

    T. Li, L. Zhang, S. Liu, and S. Shen, “Multi-modal integrated prediction and decision-making with adaptive interaction modality explorations,” arXiv preprint arXiv:2408.13742, 2024

    2024 arXiv

  23. [31]

    Surrogate safety measures for tra ffic oscillations based on empirical vehicle trajectories prior to crashes,

    Y . Wang, Z. Li, P. Liu, C. Xu, and K. Chen, “Surrogate safety measures for tra ffic oscillations based on empirical vehicle trajectories prior to crashes,” Transportation research part C: emerging technologies, vol. 161, p. 104543, 2024

    2024

  24. [32]

    Vehicle-group-based crash risk prediction and interpretation on highways,

    T. Zhu, L. Wang, Y . Feng, W. Ma, and M. Abdel-Aty, “Vehicle-group-based crash risk prediction and interpretation on highways,” arXiv preprint arXiv:2402.12415, 2024

    2024 arXiv

  25. [33]

    Spatiotemporal scene-graph embedding for autonomous vehicle collision prediction,

    A. V . Malawade, S.-Y . Yu, B. Hsu, D. Muthirayan, P. P. Khargonekar, and M. A. Al Faruque, “Spatiotemporal scene-graph embedding for autonomous vehicle collision prediction,” IEEE Internet of Things Journal , vol. 9, no. 12, pp. 9379–9388, 2022

    2022

  26. [34]

    Groupnet: Multiscale hypergraph neural networks for trajectory prediction with relational reasoning,

    C. Xu, M. Li, Z. Ni, Y . Zhang, and S. Chen, “Groupnet: Multiscale hypergraph neural networks for trajectory prediction with relational reasoning,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 6498– 6507, 2022

    2022

  27. [35]

    Hyper-relational interaction modeling in multi-modal trajectory prediction for intelligent connected vehicles in smart cites,

    Y . Lu, W. Wang, R. Bai, S. Zhou, L. Garg, A. K. Bashir, W. Jiang, and X. Hu, “Hyper-relational interaction modeling in multi-modal trajectory prediction for intelligent connected vehicles in smart cites,” Information Fusion, vol. 114, p. 102682, 2025

    2025

  28. [36]

    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

    2017

  29. [37]

    U.S. Highway 101 dataset,

    J. Colyar and J. Halkias, “U.S. Highway 101 dataset,” dataset, Federal Highway Administration, 2007. NGSIM

    2007

  30. [38]

    U.S. Highway 80 dataset,

    J. Colyar and J. Halkias, “U.S. Highway 80 dataset,” dataset, Federal Highway Administration, 2006. NGSIM

    2006

  31. [39]

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

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

    2018

  32. [40]

    Social lstm: Human trajectory prediction in crowded spaces,

    A. Alahi, K. Goel, V . Ramanathan, A. Robicquet, L. Fei-Fei, and S. Savarese, “Social lstm: Human trajectory prediction in crowded spaces,” in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 961–971, 2016

    2016

  33. [41]

    Convolutional social pooling for vehicle trajectory prediction,

    N. Deo and M. M. Trivedi, “Convolutional social pooling for vehicle trajectory prediction,” in Proceedings of the IEEE conference on computer vision and pattern recognition workshops, pp. 1468–1476, 2018

    2018

  34. [42]

    Pip: Planning-informed trajectory prediction for autonomous driving,

    H. Song, W. Ding, Y . Chen, S. Shen, M. Y . Wang, and Q. Chen, “Pip: Planning-informed trajectory prediction for autonomous driving,” inComputer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XXI 16, pp. 598–614, Springer, 2020

    2020

  35. [43]

    Grip: Graph-based interaction-aware trajectory prediction,

    X. Li, X. Ying, and M. C. Chuah, “Grip: Graph-based interaction-aware trajectory prediction,” in 2019 IEEE Intelligent Transportation Systems Conference (ITSC), pp. 3960–3966, IEEE, 2019

    2019

  36. [44]

    Intention-aware vehicle trajectory prediction based on spatial- temporal dynamic attention network for internet of vehicles,

    X. Chen, H. Zhang, F. Zhao, Y . Hu, C. Tan, and J. Yang, “Intention-aware vehicle trajectory prediction based on spatial- temporal dynamic attention network for internet of vehicles,” IEEE Transactions on Intelligent Transportation Systems , vol. 23, no. 10, pp. 19471–19483, 2022

    2022

  37. [45]

    Virtual roads, smarter safety: A digital twin framework for mixed autonomous traffic safety analysis,

    H. Zhang, X. Yue, K. Tian, S. Li, K. Wu, Z. Li, D. Lord, and Y . Zhou, “Virtual roads, smarter safety: A digital twin framework for mixed autonomous traffic safety analysis,” arXiv preprint arXiv:2504.17968, 2025. 23

    2025 arXiv

Pith tools

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