Pith. sign in

REVIEW 4 major objections 6 minor 67 references

DEMO: A Dynamics-Enhanced Learning Model for Multi-Horizon Trajectory Prediction in Autonomous Vehicles

T0 review · 4 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read DEMO claims that a two-stage architecture—a physics-based Dynamic Bicycle Model paired with a learning-based interaction stage—can predict surrounding vehicles' trajectories accurately at both short-term and long-term horizons…

desk verdict DEMO's hybrid architecture is new and its ablations are informative, but the SOTA claim is contradicted by its own tables and rests on quoted baselines. read the letter →

arxiv 2412.20784 v1 pith:QHKOKSZ5 submitted 2024-12-30 cs.RO

Chengyue Wang , Haicheng Liao , Kaiqun Zhu , Guohui Zhang , Zhenning Li This is my paper · ORCID
classification cs.RO
keywords AutonomousDrivingTrajectoryPredictionDynamics-basedModelLearning-basedDataFusionMulti-horizonDynamicBicycleConditionalVariationalAutoencoder
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 paper tries to establish that one trajectory-prediction model can be the best choice for both regimes that autonomous driving needs: reacting in the next couple of seconds and planning over a longer horizon. The proposed DEMO model splits the task into a Dynamics Learning Stage, where a Dynamic Bicycle Model and a conditional variational autoencoder estimate the vehicle's control inputs, and an Interaction Learning Stage, where a Mamba temporal encoder and cross-modal attention with HD maps model social and environmental context. On the NGSIM, MoCAD, HighD, and nuScenes datasets, DEMO reports lower errors than published state-of-the-art baselines in both short-term and long-term horizons, and an inference time of 15 ms. The authors also report that training on only half the data still beats most baselines in the short-term regime.

What carries the argument

The load-bearing mechanism is the coupling of a discrete Dynamic Bicycle Model with a Dynamic Conditional Variational Autoencoder (DynCVAE) in the Dynamics Learning Stage. The bicycle model's inverse form lets the model treat control variables, namely yaw angle, yaw rate, steering angle, and acceleration, as the quantities to estimate; the DynCVAE learns the posterior distribution over latent controls and aligns it with the prior through a KL loss, so that at inference the latent controls can be sampled without future states. A dynamics-informed loss provides supervision when no control ground truth exists. The resulting dynamic features are appended to the query and key vectors of the cross-modal attention in the Interaction Learning Stage, which is the channel through which physics informs the interaction modeling; a Mamba temporal encoder and a spatio-temporal encoder complete the interaction features, and a multi-modal decoder combines both.

What would settle it

Re-run the strongest published baselines, for example STDAN, iNATran, Q-EANet, and SeFlow, under DEMO's exact dataset splits, filtering, short-term and long-term horizon definitions, and metrics; if any baseline matches or beats DEMO's reported RMSE, minADE, or minFDE numbers on all four datasets, the paper's state-of-the-art claim is falsified.

Watch

Extended reading notes

Core claim

The paper's central claim is that the two horizons do not need to trade off against each other if the model is explicitly structured around them. In the first stage, a DynCVAE learns a distribution over latent control variables, and the discrete Dynamic Bicycle Model converts those controls into short-term trajectories and dynamic features; a dynamics-informed loss supervises this without ground-truth control labels. In the second stage, a Mamba state-space temporal encoder, a cross-modal attention mechanism that injects dynamic features into the query and key pairs along with vehicle and HD-map features, and a spatio-temporal encoder produce interaction features. A multi-modal decoder fuses both feature sets to output several trajectory hypotheses with probabilities. In the reported tables, DEMO improves short-term RMSE by 8.5% on NGSIM, by 25.7% and 44.6% on MoCAD at 1 and 2 seconds, and long-term results by 6.3–18.5% at 4 seconds and 7.3–30.7% at 5 seconds across the highway and urban datasets; on nuScenes the paper highlights improvements in minADE1 and minFDE1. The inference-time comparison reports 15 ms, faster than all compared models.

Load-bearing premise

The claim that DEMO beats state-of-the-art models assumes that the baseline numbers quoted from earlier papers were produced under evaluation conditions—dataset splits, scenario filtering, horizon definitions, and metrics—close enough to DEMO's own conditions that the differences in the tables reflect model quality rather than protocol; if the baselines were re-run under DEMO's exact protocol, the reported margins could shrink or reverse.

Editorial extensions

If this is right

  • A vehicle equipped with DEMO would not need to switch predictors when the planning horizon changes from emergency braking to lane-change strategy; the same network produces both.
  • The estimated control variables give downstream modules a physically meaningful intermediate state, not only coordinates, which could support safety checks or planning constraints.
  • Because DEMO trained on 50% of the data still beats most full-data baselines in short-term prediction, the dynamics prior appears to reduce the amount of trajectory data needed for the near-term regime.
  • The reported 15 ms inference time means the model could, in principle, issue more than sixty predictions per second, comfortably inside real-time control budgets.

Reading between the lines

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

  • Beyond the paper: the two-stage split suggests a data-efficiency curve worth testing directly—vary the training fraction continuously and measure how short-term error grows; the dynamics stage should flatten that curve relative to a purely learned baseline.
  • Beyond the paper: because the Dynamic Bicycle Model is differentiable, the same architecture could be adapted to other vehicle types such as trucks or buses by changing only physical parameters like mass, moment of inertia, axle distances, and cornering stiffness, which the paper does not explore.
  • Beyond the paper: feeding the predicted control variables to downstream planners might make the model's output auditable—a safety monitor could check whether the proposed steering or acceleration commands are feasible before execution; the paper only evaluates positional error.
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 / 6 minor

Summary. The paper proposes DEMO, a two-stage trajectory prediction architecture for autonomous driving. The Dynamics Learning Stage couples a discretized Dynamic Bicycle Model with a Dynamic Conditional Variational Autoencoder (DynCVAE) to generate short-term trajectories and dynamic features, supervised by a KL divergence term and a dynamics-informed loss. The Interaction Learning Stage then uses a Mamba-based temporal encoder, cross-modal attention with HD map features, and a spatial-temporal encoder (GRU, GCN, and Transformer) to produce interaction features, which are combined with dynamic features in a multi-modal decoder. The model is evaluated on NGSIM, MoCAD, HighD, and nuScenes with RMSE, minADE, and minFDE metrics, reporting short-term and long-term results, ablations, a 50% training-data variant, and a 15 ms inference time. The central claim is that DEMO outperforms state-of-the-art baselines in both short-term and long-term horizons while being suitable for real-time use.

Significance. The hybrid modeling idea is well motivated: using a physics-based bicycle model for the short-term stage and learned interaction modeling for the long-term stage is a reasonable design, and the ablations suggest that each component contributes to performance, especially the Dynamics Learning Stage for short-term accuracy. The 50%-data experiments are a useful indication of sample efficiency. However, the significance of the empirical comparison is currently undermined by the evaluation protocol. All baseline numbers in Tables 1–3 are quoted from prior publications rather than re-run under DEMO's protocol, and Section 4.1 does not establish a common evaluation protocol across datasets. More seriously, the paper's own tables contradict the unqualified SOTA claim: on HighD short-term, iNATran is better than DEMO, and on nuScenes, SeFlow and Q-EANet beat DEMO on minADE10 and minADE5, respectively. The architecture may still be valuable, but the claimed superiority is not established as presented.

major comments (4)
  1. [Abstract and Section 4.2, Tables 1 and 3] The abstract and conclusion state that DEMO 'outperforms state-of-the-art (SOTA) baselines in both short-term and long-term prediction horizons,' but this is contradicted by the paper's own tables. In Table 1, on HighD at the 1 s and 2 s horizons, iNATran achieves RMSE 0.04 and 0.05 while DEMO achieves 0.06 and 0.14. In Table 3, on nuScenes, SeFlow achieves minADE10 of 0.98 versus DEMO's 1.04, and Q-EANet achieves minADE5 of 1.18 versus DEMO's 1.20. The text acknowledges the HighD result in Section 4.2 but does not acknowledge the nuScenes minADE10 and minADE5 losses, and the abstract remains blanket. The claims need to be revised to be dataset-specific and horizon-specific, and the conclusion should not claim universal SOTA performance.
  2. [Section 4.1 and Section 4.2, Tables 1–3] The comparison against state-of-the-art baselines is not established because every baseline metric is quoted from its original publication and no baseline is re-implemented under DEMO's evaluation protocol. Section 4.1 defines DEMO's segmentations (3 s past / 5 s future for NGSIM, MoCAD, and HighD; 2 s past / 6 s future for nuScenes), but dataset splits, scenario filtering, map usage, the number of modes used for minADE/minFDE, and the details of evaluation all vary across the source papers. For example, DEMO's RMSE-based comparisons on NGSIM/MoCAD/HighD and minADE/minFDE comparisons on nuScenes use different protocols, yet the numbers are placed side by side in the same table. Consequently, the reported margins, such as the 8.5% improvement on NGSIM and the 25.7% improvement on MoCAD in Section 4.2, are not demonstrably real. The authors should either re-run the baselines under an identical protocol, submit to a standardized benchmark such as the nuScenes test server, or substantially temper the comparative claims.
  3. [Section 4.4 and Table 4] The inference speed comparison is not sufficiently controlled. Table 4 states that other models are evaluated on an RTX 3090 Ti GPU while DEMO is evaluated on an RTX 3090 GPU, and the comparison uses the average inference time for only 12 samples. No details are given about batching, input sizes, precision, or software implementations, and the baselines are not re-run in the same environment. The claim that DEMO is suitable for real-time applications therefore requires a more careful benchmark, or should be stated more cautiously as a single-environment observation.
  4. [Section 4.3, Tables 6–8] The ablation results are reported as single numbers without variance or statistical testing. Some of the margins in the main tables are small, such as the 0.06 difference in minADE10 on nuScenes, so it is unclear whether the improvements over baselines or the differences between ablation variants are significant. The authors should report multiple seeds with means and standard deviations, and ideally a significance test, to support the quantitative claims.
minor comments (6)
  1. [Section 3.3.1, Eq. (1)] The second row of the state equation appears to contain a typo: it reads 'v_x sin phi + v_x cos phi', but it should presumably be 'v_x sin phi + v_y cos phi' to match the standard bicycle model.
  2. [Section 3.3.2, Eq. (3)] Equation (3) writes the KL divergence as [q(z_t | X_t, X_{t+1}) || p(z_t | X_t)] without the D_KL operator; it should be written as D_KL( q(z_t | X_t, X_{t+1}) || p(z_t | X_t) ).
  3. [Section 3.3.2, Eq. (4)] Equation (4) is written as a vector equality, L_DI = X_t - hat X_t, which is not a scalar loss. It should be defined using a norm, such as the squared L2 norm, and should be described as an auxiliary regression loss.
  4. [Section 4.2, nuScenes paragraph] The text says 'in terms of minADE1 and minFDE1, we observe improvements of at least 5.4% and 1.3%, respectively,' but it does not specify which baseline this improvement is measured against. The comparison baseline should be named explicitly.
  5. [Section 4.1, Table 3] The baseline 'Physics Oracle' appears in Table 3 but is not defined or cited in the text. It should be described or referenced so that the reader can understand what it represents.
  6. [Section 4.1, Dataset Segmentations] For the NGSIM, MoCAD, and HighD datasets, HD map data is excluded and the map features are replaced with vehicle features. The paper should clarify how this substitution affects the cross-modal attention mechanism, since the architecture description in Section 3.4.2 assumes separate map and vehicle modalities.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: DEMO's dynamics loss is standard supervised fitting and the SOTA comparison is external, though protocol mismatches undermine the blanket claim.

full rationale

I find no circular step in the derivation chain. The Dynamics Learning Stage is trained with Eq. (4), L_DI = X_t - Ε(X_{t-1}, C_{t-1}), which is a supervised reconstruction loss against ground-truth trajectories; the predicted trajectory is compared with held-out future states, so the quantity being predicted is not used to define the predictor. At inference, DynCVAE samples control variables from the prior p(z|X) after KL alignment in Eq. (3), so short-term outputs are genuine predictions rather than refitted inputs. The long-term and multi-modal stages are standard learned encoders and decoders trained on external datasets. The SOTA comparisons in Tables 1-3 are benchmarked against external datasets and mostly external baselines; the two same-group baselines (BAT [37], GaVa [27]) are presented as ordinary comparison points and are not load-bearing for the architecture or the central claim. Removing them would not change the model's derivation or the external evaluation. The paper does contain a limitation statement about the inference comparison ('while the other models are evaluated on an RTX 3090 Ti GPU [42], DEMO is tested on the RTX 3090 GPU'), and the blanket 'outperforms SOTA' claim is internally contradicted by Table 1 (HighD short-term iNATran) and Table 3 (nuScenes minADE10 SeFlow/Q-EANet), but these are correctness and external-validity concerns, not circularity. No uniqueness theorem, ansatz-smuggling citation, or definitional equivalence is present. Therefore the circularity score is 0.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The central empirical claim rests on a small number of undisclosed implementation choices (bicycle model parameters, network hyperparameters) and two strong domain assumptions (model validity and baseline comparability). No new physical entities are introduced.

free parameters (2)
  • Vehicle physical parameters of Dynamic Bicycle Model (m, Iz, lf, lr, kf, kr)
    These parameters appear in Eqs. (1)-(2) and are required to compute state transitions, but the paper never states how they are set for each dataset. They are treated as known constants without disclosure, which is a free choice that can substantially affect short-term dynamics.
  • Network hyperparameters (latent dim, attention blocks N, hidden sizes, loss weights)
    Section 4.1 gives only optimizer and learning rate; the sizes of the DynCVAE, Mamba, cross-modal attention, GCN layers, and the weighting of L_KL, L_DI, L_CE, L_AC are unspecified. These choices are tuned during development and are needed for reproduction.
assumptions (3)
  • domain assumption Dynamic Bicycle Model (Eq. 1) adequately represents vehicle motion for all scenarios in NGSIM, MoCAD, HighD, and nuScenes.
    Section 3.3.1 assumes the model captures essential lateral and longitudinal dynamics, but no validation against real vehicle data is provided. If the model mismatches certain vehicle types (e.g., trucks, motorcycles), the dynamics stage could be biased.
  • domain assumption Quoted baseline metrics from different papers are directly comparable without re-implementation.
    Section 4.2 imports baseline numbers from heterogeneous sources with different dataset splits, preprocessing, and evaluation pipelines. This is the load-bearing premise for the SOTA claim.
  • domain assumption The 50% subset results (DEMO (50%)) are representative; the split method is not described.
    Section 4.2 reports DEMO (50%) as a robustness check but does not state which 50% of data was used or whether the split was random and repeated. This makes the efficiency claim hard to interpret.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DEMO: A Dynamics-Enhanced Learning Model for Multi-Horizon Trajectory Prediction in Autonomous Vehicles." pith.science (2026). https://pith.science/paper/QHKOKSZ5

@misc{pith2026241220784,
  author       = {Pith},
  title        = {Pith review of: DEMO: A Dynamics-Enhanced Learning Model for Multi-Horizon Trajectory Prediction in Autonomous Vehicles},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QHKOKSZ5}},
  note         = {Machine review of arXiv:2412.20784}
}
read the original abstract

Autonomous vehicles (AVs) rely on accurate trajectory prediction of surrounding vehicles to ensure the safety of both passengers and other road users. Trajectory prediction spans both short-term and long-term horizons, each requiring distinct considerations: short-term predictions rely on accurately capturing the vehicle's dynamics, while long-term predictions rely on accurately modeling the interaction patterns within the environment. However current approaches, either physics-based or learning-based models, always ignore these distinct considerations, making them struggle to find the optimal prediction for both short-term and long-term horizon. In this paper, we introduce the Dynamics-Enhanced Learning MOdel (DEMO), a novel approach that combines a physics-based Vehicle Dynamics Model with advanced deep learning algorithms. DEMO employs a two-stage architecture, featuring a Dynamics Learning Stage and an Interaction Learning Stage, where the former stage focuses on capturing vehicle motion dynamics and the latter focuses on modeling interaction. By capitalizing on the respective strengths of both methods, DEMO facilitates multi-horizon predictions for future trajectories. Experimental results on the Next Generation Simulation (NGSIM), Macau Connected Autonomous Driving (MoCAD), Highway Drone (HighD), and nuScenes datasets demonstrate that DEMO outperforms state-of-the-art (SOTA) baselines in both short-term and long-term prediction horizons.

Figures

Figures reproduced from arXiv: 2412.20784 by the authors.

Figure 1
Figure 1. Comparison between the traditional Physics-based model (a), Learning-based model (b), and our proposed DEMO (c). Unlike previous approaches, DEMO employs a two-stage architecture for hierarchical prediction, integrating a Dynamics Learning Stage with an Interaction Learning Stage. social dynamics over time, which is essential for strategic decision-making that balances immediate reactivity with long-term objectives … view at source ↗
Figure 2
Figure 2. Overall architecture of DEMO. In the Dynamics Learning Stage, DEMO takes the historical trajectory as input, processes it through the DynCVAE and the Dynamic Bicycle Model, and ultimately generates the dynamic features and short-term trajectory. In the subsequent Dependencies Learning Stage, the model leverages the enhanced output from the previous stage, along with HD map data, to model the interaction present in t… view at source ↗
Figure 3
Figure 3. Illustration of the Cross-modal Fusion. Panel (a) presents the pipeline of the Cross-modal Fusion process, while Panel (b) provides a detailed illustration of the mechanism of the Cross-modal Attention Head. Regression Head. Following the cross-modal attention, we integrate a regression head at the top of this component. The high-dimensional hidden embedding 𝑒 is fed into this regression head, which consists of a 4… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Qualitative comparison of various models on nuScenes dataset. We select some representative scenarios including T-intersection (a), roundabout (b), interchange (c), and intersection (d). Panels (a), (b), and (c) visualize the most probable predictions of each model alo…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

67 extracted references · 58 canonical work pages

  1. [1]

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

    work page 2025

  2. [2]

    Z. Lan, Y. Ren, H. Yu, L. Liu, Z. Li, Y. Wang, Z. Cui, Hi-scl: Fighting long-tailed challenges in trajectory prediction with hierarchical wave-semantic contrastive learning, Transportation Research Part C: Emerging Technologies 165 (2024) 104735

    work page 2024

  3. [3]

    Z. Li, Z. Chen, Y. Li, C. Xu, Context-aware trajectory prediction for autonomous driving in heterogeneous environments, Computer-Aided Civil and Infrastructure Engineering 39 (2024) 120–135

    work page 2024

  4. [4]

    X.Chen,H.Zhang,F.Zhao,Y.Hu,C.Tan,J.Yang, Intention-awarevehicletrajectorypredictionbasedonspatial-temporaldynamicattention network for internet of vehicles, IEEE Transactions on Intelligent Transportation Systems 23 (2022) 19471–19483

    work page 2022

  5. [5]

    R.Chandra,T.Guan,S.Panuganti,T.Mittal,U.Bhattacharya,A.Bera,D.Manocha, Forecastingtrajectoryandbehaviorofroad-agentsusing spectral clustering in graph-lstms, IEEE Robotics and Automation Letters 5 (2020) 4882–4890

    work page 2020

  6. [6]

    C.Anderson,R.Vasudevan,M.Johnson-Roberson, Akinematicmodelfortrajectorypredictioningeneralhighwayscenarios, IEEERobotics and Automation Letters 6 (2021) 6757–6764

    work page 2021

  7. [7]

    G.Xie,H.Gao,L.Qian,B.Huang,K.Li,J.Wang, Vehicletrajectorypredictionbyintegratingphysics-andmaneuver-basedapproachesusing interactive multiple models, IEEE Transactions on Industrial Electronics 65 (2017) 5999–6008

    work page 2017

  8. [8]

    X. Mo, Y. Xing, C. Lv, Heterogeneous graph social pooling for interaction-aware vehicle trajectory prediction, Transportation Research Part E: Logistics and Transportation Review 191 (2024) 103748

    work page 2024

Show all 67 references
  1. [9]

    Y. Ren, Z. Lan, L. Liu, H. Yu, Emsin: Enhanced multi-stream interaction network for vehicle trajectory prediction, IEEE Transactions on Fuzzy Systems (2024)

    2024

  2. [10]

    Wang et al.:Preprint submitted to Elsevier Page 14 of 16 Information Fusion

    C.Feng,H.Zhou,H.Lin,Z.Zhang,Z.Xu,C.Zhang,B.Zhou,S.Shen, Macformer:Map-agentcoupledtransformerforreal-timeandrobust trajectory prediction, IEEE Robotics and Automation Letters (2023). Wang et al.:Preprint submitted to Elsevier Page 14 of 16 Information Fusion

    2023

  3. [11]

    M. Geng, J. Li, C. Li, N. Xie, X. Chen, D.-H. Lee, Adaptive and simultaneous trajectory prediction for heterogeneous agents via transferable hierarchical transformer network, IEEE Transactions on Intelligent Transportation Systems 24 (2023) 11479–11492

    2023

  4. [12]

    Y.Huang,J.Du,Z.Yang,Z.Zhou,L.Zhang,H.Chen, Asurveyontrajectory-predictionmethodsforautonomousdriving, IEEETransactions on Intelligent Vehicles 7 (2022) 652–674

    2022

  5. [13]

    Houenou, P

    A. Houenou, P. Bonnifait, V. Cherfaoui, W. Yao, Vehicle trajectory prediction based on motion model and maneuver recognition, in: 2013 IEEE/RSJ international conference on intelligent robots and systems, IEEE, 2013, pp. 4363–4369

    2013

  6. [14]

    Marzbani, H

    H. Marzbani, H. Khayyam, Ð. V. Quoc, R. N. Jazar, Autonomous vehicles: Autodriver algorithm and vehicle dynamics, IEEE Transactions on Vehicular Technology 68 (2019) 3201–3211

    2019

  7. [15]

    Y.Liu,J.Zhou,D.Tian,Z.Sheng,X.Duan,G.Qu,V.C.Leung, Jointcommunicationandcomputationresourceschedulingofauav-assisted mobile edge computing system for platooning vehicles, IEEE Transactions on Intelligent Transportation Systems 23 (2021) 8435–8450

    2021

  8. [16]

    Zhang, Y

    Y. Zhang, Y. Lin, Y. Qin, M. Dong, L. Gao, E. Hashemi, A new adaptive cruise control considering crash avoidance for intelligent vehicle, IEEE Transactions on Industrial Electronics 71 (2023) 688–696

    2023

  9. [17]

    S. Yang, Y. Lu, S. Li, An overview on vehicle dynamics, International Journal of Dynamics and Control 1 (2013) 385–395

    2013

  10. [18]

    Y. Liu, J. Zhou, D. Tian, Z. Sheng, X. Duan, G. Qu, D. Zhao, Joint optimization of resource scheduling and mobility for uav-assisted vehicle platoons, in: 2021 IEEE 94th Vehicular Technology Conference (VTC2021-Fall), IEEE, 2021, pp. 1–5

    2021

  11. [19]

    M.Liu,K.Chour,S.Rathinam,S.Darbha,Lateralcontrolofanautonomousandconnectedfollowingvehiclewithlimitedpreviewinformation, IEEE Transactions on Intelligent Vehicles 6 (2020) 406–418

    2020

  12. [20]

    Q. Ge, Q. Sun, S. E. Li, S. Zheng, W. Wu, X. Chen, Numerically stable dynamic bicycle model for discrete-time control, in: 2021 IEEE Intelligent Vehicles Symposium Workshops (IV Workshops), IEEE, 2021, pp. 128–134

    2021

  13. [21]

    G.Bellegarda,Q.Nguyen, Dynamicvehicledriftingwithnonlinearmpcandafusedkinematic-dynamicbicyclemodel, IEEEControlSystems Letters 6 (2021) 1958–1963

    2021

  14. [22]

    H. Liao, Z. Li, C. Wang, B. Wang, H. Kong, Y. Guan, G. Li, Z. Cui, C. Xu, A cognitive-driven trajectory prediction model for autonomous driving in mixed autonomy environment, International Joint Conference On Artificial Intelligence (IJCAI 2024) (2024)

    2024

  15. [23]

    15065–15076

    D.Park,J.Jeong,S.-H.Yoon,J.Jeong,K.-J.Yoon, T4p:Test-timetrainingoftrajectorypredictionviamaskedautoencoderandactor-specific token memory, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 15065–15076

    2024

  16. [24]

    H. Liao, X. Li, Y. Li, H. Kong, C. Wang, B. Wang, Y. Guan, K. Tam, Z. Li, C. Xu, Characterized diffusion and spatial-temporal interaction network for trajectory prediction in autonomous driving, International Joint Conference On Artificial Intelligence (IJCAI 2024) (2024)

    2024

  17. [25]

    Zhang, X

    K. Zhang, X. Feng, L. Wu, Z. He, Trajectory prediction for autonomous driving using spatial-temporal graph attention transformer, IEEE Transactions on Intelligent Transportation Systems 23 (2022) 22343–22353

    2022

  18. [26]

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

    2023

  19. [27]

    H. Liao, S. Liu, Y. Li, Z. Li, C. Wang, B. Wang, Y. Guan, C. Xu, Human observation-inspired trajectory prediction for autonomous driving in mixed-autonomy traffic environments, IEEE International Conference on Robotics and Automation (ICRA) (2024)

    2024

  20. [28]

    X. Lin, T. Liang, J. Lai, J.-F. Hu, Progressive pretext task learning for human trajectory prediction, in: ECCV, 2024

    2024

  21. [29]

    Z.Lan,L.Liu,B.Fan,Y.Lv,Y.Ren,Z.Cui, Traj-llm:Anewexplorationforempoweringtrajectorypredictionwithpre-trainedlargelanguage models, IEEE Transactions on Intelligent Vehicles (2024)

    2024

  22. [30]

    I. Bae, J. Lee, H.-G. Jeon, Can language beat numerical regression? language-based multimodal trajectory prediction, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 753–766

    2024

  23. [31]

    Xu, J.-B

    P. Xu, J.-B. Hayet, I. Karamouzas, Context-aware timewise vaes for real-time vehicle trajectory prediction, IEEE Robotics and Automation Letters (2023)

    2023

  24. [32]

    Aoki, State space modeling of time series, Springer Science & Business Media, 2013

    M. Aoki, State space modeling of time series, Springer Science & Business Media, 2013

    2013

  25. [33]

    Y.N.Dauphin,A.Fan,M.Auli,D.Grangier, Languagemodelingwithgatedconvolutionalnetworks, in:Internationalconferenceonmachine learning, PMLR, 2017, pp. 933–941

    2017

  26. [34]

    Hendrycks, K

    D. Hendrycks, K. Gimpel, Gaussian error linear units (gelus), arXiv preprint arXiv:1606.08415 (2016)

    2016 arXiv

  27. [35]

    N. Deo, M. M. Trivedi, Convolutional social pooling for vehicle trajectory prediction, in: Proceedings of the IEEE CVPR workshops, 2018, pp. 1468–1476

    2018

  28. [36]

    Krajewski,J

    R. Krajewski,J. Bock,L. Kloeker, L.Eckstein, The highd dataset:A drone datasetof naturalisticvehicle trajectories ongerman highwaysfor validation of highly automated driving systems, in: 2018 21st International Conference on Intelligent Transportation Systems (ITSC), 2018, pp....

    2018

  29. [37]

    H. Liao, Z. Li, H. Shen, W. Zeng, D. Liao, G. Li, C. Xu, Bat: Behavior-aware human-like trajectory prediction for autonomous driving, in: Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, 2024, pp. 10332–10340

    2024

  30. [38]

    H.Caesar, V.Bankiti, A.H. Lang,S. Vora,V. E.Liong, Q.Xu, A.Krishnan, Y.Pan, G.Baldan, O.Beijbom, nuscenes: Amultimodaldataset for autonomous driving, in: Proceedings of the IEEE/CVF CVPR, 2020, pp. 11621–11631

    2020

  31. [39]

    Loshchilov, F

    I. Loshchilov, F. Hutter, Decoupled weight decay regularization, arXiv preprint arXiv:1711.05101 (2017)

    2017 arXiv

  32. [40]

    Loshchilov, F

    I. Loshchilov, F. Hutter, Sgdr: Stochastic gradient descent with warm restarts, in: International Conference on Learning Representations, 2022

    2022

  33. [41]

    H. Liao, Z. Li, C. Wang, H. Shen, B. Wang, D. Liao, G. Li, C. Xu, Mftraj: Map-free, behavior-driven trajectory prediction for autonomous driving, International Joint Conference On Artificial Intelligence (IJCAI 2024) (2024)

    2024

  34. [42]

    S. Moon, H. Woo, H. Park, H. Jung, R. Mahjourian, H.-g. Chi, H. Lim, S. Kim, J. Kim, Visiontrap: Vision-augmented trajectory prediction guided by textual descriptions, in: European Conference on Computer Vision, Springer, 2025, pp. 361–379

    2025

  35. [43]

    H. Liao, Y. Li, Z. Li, C. Wang, C. Tian, Y. Huang, Z. Bian, K. Zhu, G. Li, Z. Pu, et al., Less is more: Efficient brain-inspired learning for autonomous driving trajectory prediction, European Conference on Artificial Intelligence (ECAI 2024) (2024). Wang et al.:Preprint submi...

    2024

  36. [44]

    Gupta, J

    A. Gupta, J. Johnson, L. Fei-Fei, S. Savarese, A. Alahi, Social gan: Socially acceptable trajectories with generative adversarial networks, in: Proceedings of the IEEE CVPR, 2018, pp. 2255–2264

    2018

  37. [45]

    K.Messaoud,I.Yahiaoui,A.Verroust-Blondet,F.Nashashibi, Non-localsocialpoolingforvehicletrajectoryprediction, in:IEEEIntelligent Vehicles Symposium (IV), IEEE, 2019, pp. 975–980

    2019

  38. [46]

    X. Xie, C. Zhang, Y. Zhu, Y. N. Wu, S.-C. Zhu, Congestion-aware multi-agent trajectory prediction for collision avoidance, in: 2021 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2021, pp. 13693–13700

    2021

  39. [47]

    X. Chen, H. Zhang, F. Zhao, Y. Cai, H. Wang, Q. Ye, Vehicle trajectory prediction based on intention-aware non-autoregressive transformer with multi-attention learning for internet of vehicles, IEEE Transactions on Instrumentation and Measurement 71 (2022) 1–12

    2022

  40. [48]

    K.Messaoud,I.Yahiaoui,A.Verroust-Blondet,F.Nashashibi, Attentionbasedvehicletrajectoryprediction, IEEETransactionsonIntelligent Vehicles 6 (2021) 175–185

    2021

  41. [49]

    P. Cong, Y. Xiao, X. Wan, M. Deng, J. Li, X. Zhang, Dacr-amtp: Adaptive multi-modal vehicle trajectory prediction for dynamic drivable areas based on collision risk, IEEE Transactions on Intelligent Vehicles (2023)

    2023

  42. [50]

    R. Wang, S. Wang, H. Yan, X. Wang, Wsip: Wave superposition inspired pooling for dynamic interactions-aware trajectory prediction, in: Proceedings of the AAAI Conference on Artificial Intelligence, 2023, pp. 4685–4692

    2023

  43. [51]

    Y. Chai, B. Sapp, M. Bansal, D. Anguelov, Multipath: Multiple probabilistic anchor trajectory hypotheses for behavior prediction, in: Conference on Robot Learning, PMLR, 2020, pp. 86–99

    2020

  44. [52]

    Phan-Minh, E

    T. Phan-Minh, E. C. Grigore, F. A. Boulton, O. Beijbom, E. M. Wolff, Covernet: Multimodal behavior prediction using trajectory sets, in: Proceedings of the IEEE/CVF CVPR, 2020, pp. 14074–14083

    2020

  45. [53]

    Y. Yuan, K. Kitani, Dlow: Diversifying latent flows for diverse human motion prediction, in: Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part IX 16, Springer, 2020, pp. 346–364

    2020

  46. [54]

    T.Salzmann,B.Ivanovic,P.Chakravarty,M.Pavone, Trajectron++:Dynamically-feasibletrajectoryforecastingwithheterogeneousdata, in: Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XVIII 16, Springer, 2020, pp. 683–700

    2020

  47. [55]

    13279–13288

    Y.J.Ma,J.P.Inala,D.Jayaraman,O.Bastani, Likelihood-baseddiversesamplingfortrajectoryforecasting, in:ProceedingsoftheIEEE/CVF International Conference on Computer Vision, 2021, pp. 13279–13288

    2021

  48. [56]

    Messaoud, N

    K. Messaoud, N. Deo, M. M. Trivedi, F. Nashashibi, Trajectory prediction for autonomous driving based on multi-head attention with joint agent-map representation, in: 2021 IEEE Intelligent Vehicles Symposium (IV), IEEE, 2021, pp. 165–170

    2021

  49. [57]

    B. Kim, S. H. Park, S. Lee, E. Khoshimjonov, D. Kum, J. Kim, J. S. Kim, J. W. Choi, Lapred: Lane-aware prediction of multi-modal future trajectoriesofdynamicagents, in:ProceedingsoftheIEEE/CVFConferenceonComputerVisionandPatternRecognition,2021,pp.14636– 14645

    2021

  50. [58]

    Zhong, Y

    Z. Zhong, Y. Luo, W. Liang, Stgm: Vehicle trajectory prediction based on generative model for spatial-temporal features, IEEE Transactions on Intelligent Transportation Systems 23 (2022) 18785–18793

    2022

  51. [59]

    Gilles, S

    T. Gilles, S. Sabatini, D. Tsishkou, B. Stanciulescu, F. Moutarde, Gohome: Graph-oriented heatmap output for future motion estimation, in: 2022 international conference on robotics and automation (ICRA), IEEE, 2022, pp. 9107–9114

    2022

  52. [60]

    N. Deo, E. Wolff, O. Beijbom, Multimodal trajectory prediction conditioned on lane-graph traversals, in: Conference on Robot Learning, PMLR, 2022, pp. 203–212

    2022

  53. [61]

    J.Chen,Z.Wang,J.Wang,B.Cai, Q-eanet:Implicitsocialmodelingfortrajectorypredictionviaexperience-anchoredqueries, IETIntelligent Transport Systems 18 (2024) 1004–1015

    2024

  54. [62]

    15226–15237

    Y.Xu,Y.Fu,Adaptingtolengthshift:Flexilengthnetworkfortrajectoryprediction,in:ProceedingsoftheIEEE/CVFConferenceonComputer Vision and Pattern Recognition, 2024, pp. 15226–15237

    2024

  55. [63]

    C. Wong, B. Xia, Z. Zou, Y. Wang, X. You, Socialcircle: Learning the angle-based social interaction representation for pedestrian trajectory prediction, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 19005–19015

    2024

  56. [64]

    Zhang, Y

    Q. Zhang, Y. Yang, P. Li, O. Andersson, P. Jensfelt, Seflow: A self-supervised scene flow method in autonomous driving, in: European Conference on Computer Vision, Springer, 2025, pp. 353–369

    2025

  57. [65]

    N. Deo, M. M. Trivedi, Trajectory forecasts in unknown environments conditioned on grid-based plans, arXiv preprint arXiv:2001.00735 (2020)

    2020 arXiv

  58. [66]

    Y. Yuan, X. Weng, Y. Ou, K. M. Kitani, Agentformer: Agent-aware transformers for socio-temporal multi-agent forecasting, in: Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 9813–9823

    2021

  59. [67]

    M. Liu, H. Cheng, L. Chen, H. Broszio, J. Li, R. Zhao, M. Sester, M. Y. Yang, Laformer: Trajectory prediction for autonomous driving with lane-aware scene constraints, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 2039–2049. W...

    2024

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.