Pith. sign in

REVIEW 5 major objections 6 minor 56 references

Characterized Diffusion Networks for Enhanced Autonomous Driving Trajectory Prediction

T0 review · 5 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read A model coupling a characterized diffusion denoiser with a spatial-temporal attention network achieves state-of-the-art 5-second trajectory prediction on NGSIM, HighD, and MoCAD, with an RMSE of 2.85 meters on NGSIM.

desk verdict The paper's core formulation is not implemented as described; the benchmark claims are unsupported, so this is not ready for review. read the letter →

arxiv 2411.16457 v1 pith:JANXQJNH submitted 2024-11-25 cs.RO cs.AI

Haoming Li This is my paper · ORCID
classification cs.ROcs.AI
keywords trajectorypredictionautonomousdrivingdiffusionmodelsspatial-temporalattentionuncertaintyestimationNGSIMHighDMoCAD
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 proposes CDSTraj, a generative model for predicting the future paths of surrounding vehicles in autonomous driving. The model combines a Characterized Diffusion Module, which iteratively removes noise from candidate trajectories, with a Spatial-Temporal Interaction Network that attends to both the positions and the evolving behavior of neighboring agents. The authors claim this coupling addresses two shortcomings of prior work: focusing only on the target agent's uncertainty while ignoring the wider scene, and treating spatial interactions as static rather than time-varying. They report substantial gains, including a 5-second RMSE of 2.85 meters on NGSIM versus 3.40 meters for the next-best model, with similar improvements on HighD and MoCAD. If these results hold, the framework would offer an immediately usable accuracy boost for motion planning in dense and mixed-traffic settings.

What carries the argument

The central machinery is a two-part generative architecture. The Characterized Diffusion Module runs a forward noising process $C_\delta = f_{\text{diff}}(C_{\delta-1})$ and a reverse denoising loop that conditions on historical states and adaptively scales updates via $\alpha_\delta$ and $\bar{\alpha}_\delta$ in the rule $\hat{C}_\delta = \frac{1}{\sqrt{\alpha_\delta}}(\hat{C}_{\delta+1} - \frac{1-\alpha_\delta}{\sqrt{1-\bar{\alpha}_\delta}} \hat{\epsilon}_\delta) + \sqrt{\frac{1-\alpha_\delta}{\alpha_\delta}} z$. The Spatial-Temporal Interaction Network computes multi-head attention weights $\omega = \text{softmax}(QK^T/\sqrt{d})$ over agents, then fuses spatial attention output $\Upsilon$ with temporal recurrent states through a gate $S = H_a \odot H_g$. An LSTM decoder turns the fused representation into predicted positions. The interaction between the two modules is the defining innovation: the diffusion module generates the neighbor futures $Y_i$ that the interaction network then uses to refine the target trajectory.

What would settle it

Run CDSTraj on NGSIM with $Y_i$ produced by the model's own decoder instead of ground-truth future frames and measure the 5-second RMSE; if the error rises from 2.85 m to roughly the 3.4-3.7 m range of the strongest baselines, the original result relied on oracle neighbor futures.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central claim is that trajectory prediction accuracy improves when uncertainty is modeled by a diffusion process over the whole scene and when agent interactions are encoded jointly over space and time. Concretely, CDSTraj defines the target trajectory as $Y_0 = \Phi(X_0, X_i, Y_i)$ for all neighbors $i$, meaning the prediction explicitly consumes the neighboring agents' future trajectories $Y_i$ alongside their histories. A forward diffusion process adds controlled noise to candidate trajectories and a reverse process denoises them using the historical context, with step-specific parameters $\alpha_\delta$ and $\bar{\alpha}_\delta$ that adapt the update scale. The spatial-temporal module uses multi-head attention for pairwise interactions and a gated fusion $S = H_a \odot H_g$ to combine the two dimensions. In experiments, the model reports RMSE values of 0.36, 0.86, 1.36, 2.02, and 2.85 meters at horizons 1 through 5 seconds on NGSIM, and similar improvements on HighD and MoCAD, with the gap over the strongest baselines growing at longer horizons.

Load-bearing premise

The model's defining equation $Y_0 = \Phi(X_0, X_i, Y_i)$ requires the future trajectories of neighboring agents, and the paper never states whether those come from ground truth, from the model itself, or from an external predictor, so the reported accuracy gains may depend on information that is not available at inference time.

Editorial extensions

If this is right

  • If the reported numbers hold, CDSTraj would be the most accurate published predictor on NGSIM, HighD, and MoCAD at horizons up to 5 seconds, with the largest margins in long-term predictions.
  • The two-stage training schedule (MSE followed by NLL) allows the model to output not only point trajectories but also per-step uncertainty parameters ($\sigma^x_t$, $\sigma^y_t$, $\rho^{xy}_t$), which can quantify confidence in each predicted position.
  • The model's performance on MoCAD, which has left-hand traffic and varied urban conditions, suggests the architecture transfers across driving conventions without reconfiguration.
  • The ablation study attributes most of the gain to the characterized diffusion module and confidence feature fusion, since removing either raises NGSIM RMSE from 2.85 m to roughly 3.0-3.1 m.

Reading between the lines

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

  • The paper never specifies whether $Y_i$ in Eq. (1) come from ground-truth future labels, from the model's own predictions, or from an external predictor; if they are ground-truth, the reported benchmark gains would reflect oracle information, and a fair comparison would require ablating this input.
  • If $Y_i$ are model outputs, then CDSTraj is effectively a joint prediction model for all agents, and its inference cost and error accumulation across agents should be analyzed; the paper does not provide this analysis.
  • The uncertainty parameters ($\sigma$ and $\rho$) produced by the NLL stage could support risk-aware planning or collision-avoidance modules, an application the paper mentions in general terms but does not evaluate.
  • A natural extension is to replace the simple LSTM decoder with a trajectory-sampling head that outputs a full distribution, which would let the diffusion model's diversity be evaluated with metrics like minADE and miss rate rather than mean RMSE.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 6 minor

Summary. The paper introduces CDSTraj, a trajectory prediction model for autonomous driving that combines a 'Characterized Diffusion Module' with a 'Spatial-Temporal Interaction Network.' The central idea, expressed in Eq. (1), is to predict a target agent's future trajectory from the agent's and neighbors' histories and from the predicted future trajectories of neighboring agents. The authors report experiments on NGSIM, HighD, and MoCAD and claim substantial improvements over existing methods; Table 1 reports NGSIM RMSE values of 0.36 m at 1 s and 2.85 m at 5 s. An ablation study over five components is also provided. The main unresolved issue is that the provenance of the neighbor-future term Yi is never specified, and the evidence for the headline claims is incomplete.

Significance. If the reported performance were reproducible, the 5 s NGSIM RMSE of 2.85 m would be a competitive result, and the combination of diffusion with spatial-temporal attention would be worth studying. The manuscript's strengths are the breadth of the baseline comparison in Table 1 and the effort to ablate the proposed components. However, the key formulation in Eq. (1) is ambiguous in a way that directly affects the benchmark numbers, no HighD result table is provided, and no error bars, seeds, or evaluation protocol details are given. As a consequence, the claimed state-of-the-art performance cannot be assessed or reproduced.

major comments (5)
  1. [Section 3, Eq. (1); Section 4.1, Eq. (5)] The central formulation Y0 = Phi(X0, Xi, Yi) makes the prediction of the target agent depend on the future trajectories Yi of neighboring agents, but the manuscript never states where Yi comes from at inference. In Section 4.1, the reverse diffusion in Eq. (5) conditions only on X0 and Xi, and the encoding and decoding stages in Sections 4.2-4.3 similarly do not consume Yi, so the equation is not implemented as written. If Yi are ground-truth future trajectories, the NGSIM and MoCAD results are inflated by information leakage; if Yi are the model's own outputs, the method is a closed-loop joint predictor whose error accumulation is not analyzed; if Yi is only notational, the stated key innovation is absent. The authors should specify the inference-time source of Yi and provide an ablation with and without this term.
  2. [Section 5.3] Section 5.3 reports large improvements on HighD (43%-70% for short-term predictions and 62%-78% for long-term predictions), but no HighD results table or numerical values appear anywhere in the manuscript. The MoCAD results are presented in Table 2 and the NGSIM results in Table 1, but the HighD claims are unverifiable. Either the HighD table should be added or the corresponding claims should be removed.
  3. [Section 5, Tables 1-3] No evaluation protocol is given: the manuscript does not state the train/validation/test split, the number of samples, the number of random seeds, or standard deviations for any reported RMSE value. In addition, the diffusion process produces K samples in Eq. (4), but the paper never explains how the K trajectories are converted into the scalar RMSE values in Tables 1-3, such as best-of-K, mean, or winner-take-all. Without this information, the comparison with single-shot baselines cannot be interpreted.
  4. [Section 5.4, Table 3] The ablation study reports a single RMSE value per configuration without stating the prediction horizon, and the table has no error bars, so the reader cannot judge whether the differences between configurations are significant. The 'Confidence feature fusion' component, which is described as important in Section 5.4, is not defined anywhere in the methodology and is not listed among the ablated components in Table 3.
  5. [Abstract and Section 5.3, Table 1] The quantitative claims in the abstract and text are not consistent with the table. The abstract says the model 'significantly outperforms' existing methods, but Table 1 shows BAT has lower RMSE at 1 s (0.23 vs 0.36) and at 2 s (0.81 vs 0.86). The text claims a 29% improvement over WSiP, but the table values (4.34 vs 2.85) imply a 34% improvement. The claims should be recomputed and restated to reflect the actual table entries.
minor comments (6)
  1. [Section 1 and Section 2] Several citations are broken: Section 1 contains '[37; ?]' and Section 2's 'Xi et al. (2024)' has no corresponding numbered reference.
  2. [Section 6] The conclusion ends with the incomplete phrase 'the integra...' and should be completed.
  3. [Figures] The four figures listed at the end of the manuscript (cdffusion1.png, exp111.png, head.png, network.png) are never referenced in the text and have no captions in the submitted text.
  4. [Section 5.2, Eq. (18)] The NLL loss in Eq. (18) introduces alpha, rho_xy, and P_t without defining them or explaining how the NLL loss is combined with the MSE objective in the two-stage training procedure.
  5. [Table 1] The table caption states that bold and underlined values represent the best and second-best performance, but no underlining appears in the rendered table.
  6. [Section 5.1] The MoCAD dataset is described but no citation is provided for it, despite being one of the three evaluation benchmarks.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the implemented equations are not derived from the paper's own outputs, and the Yi term in Eq. (1), while underspecified, is a correctness/leakage concern rather than a circular reduction.

full rationale

I could not exhibit a circular step by the paper's own equations. Eq. (1) writes the target future Y0 as a function of neighbor predicted futures Yi, but this is a conditional two-stage formulation rather than an identity: Yi is treated as an auxiliary output of the Characterized Diffusion Module, not as Y0 itself. The reverse-diffusion equations (5)-(6) condition on historical states X0, Xi; the decoder (16) generates positions from the fused representation S. None of these equations re-insert Y0 into its own input in a way that makes the output equal to the input by construction. The missing provenance of Yi (whether it is produced by the diffusion module, by ground truth, or by an external predictor) is a serious reproducibility and potential future-leakage risk, but the paper does not fit Yi to Y0 or define Y0 in terms of Y0, so it is not a circularity finding under the stated rules. The DDPM-style update in Eq. (7) is standard diffusion machinery cited in Related Work; relabeling it as 'Characterized Diffusion' is a novelty concern, not circularity. No load-bearing self-citation appears: the H. Liao et al. references are not the present author's own, and the baselines are external. I therefore set score 0.

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

The central claim rests on an unobservable input (neighbor futures), an unspecified diffusion training objective, and unreported hyperparameters. The paper introduces two vaguely named modules but no new measurable entity, so the ledger is mostly composed of domain assumptions and hidden parameter choices.

free parameters (3)
  • Diffusion step-size parameters alpha_delta and bar_alpha_delta = unspecified
    Eqs. (6)-(7) use step-specific scaling that controls the denoising update; the paper does not state whether they are learned or fixed, and the final trajectories depend on them.
  • NLL loss weighting alpha = unspecified
    Introduced in Eq. (18); the value is not reported and would change the uncertainty calibration and predictions.
  • Diffusion steps Gamma and sample count K = unspecified
    Eqs. (3) and (5) depend on Gamma, and Eq. (4) on K; neither is given, and both affect output diversity and accuracy.
assumptions (3)
  • ad hoc to paper The target agent's future trajectory is predictable from the histories X0, Xi plus the neighboring agents' future trajectories Yi.
    This is Eq. (1) in Section 3; Yi is a future quantity that is not observable, and the paper never resolves this dependency.
  • domain assumption Standard DDPM reverse sampling (Eqs. 6-7) is an appropriate generative model for vehicle trajectories.
    Imported from Ho et al. 2020 without any validation or adaptation for trajectory data; the training objective for f_epsilon is missing.
  • domain assumption RMSE on NGSIM, HighD, and MoCAD is the right metric for comparing trajectory prediction quality.
    The paper reports only RMSE and omits metrics such as minADE, minFDE, and collision rate, which are standard in the trajectory prediction community.
invented entities (2)
  • Characterized Diffusion Module
    purpose: Claimed to reduce uncertainty in neighboring agents' future trajectories and improve target-agent prediction.
    The term 'characterized' is never defined; the equations are the standard DDPM update, so the module has no independent falsifiable handle.
  • Confidence feature fusion
    purpose: Claimed in the ablation section to contribute to accuracy via uncertainty features.
    This component is not described in the methodology and appears only in the Table 3 caption, with no definition or separate evidence.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Characterized Diffusion Networks for Enhanced Autonomous Driving Trajectory Prediction." pith.science (2026). https://pith.science/paper/JANXQJNH

@misc{pith2026241116457,
  author       = {Pith},
  title        = {Pith review of: Characterized Diffusion Networks for Enhanced Autonomous Driving Trajectory Prediction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JANXQJNH}},
  note         = {Machine review of arXiv:2411.16457}
}
read the original abstract

In this paper, we present a novel trajectory prediction model for autonomous driving, combining a Characterized Diffusion Module and a Spatial-Temporal Interaction Network to address the challenges posed by dynamic and heterogeneous traffic environments. Our model enhances the accuracy and reliability of trajectory predictions by incorporating uncertainty estimation and complex agent interactions. Through extensive experimentation on public datasets such as NGSIM, HighD, and MoCAD, our model significantly outperforms existing state-of-the-art methods. We demonstrate its ability to capture the underlying spatial-temporal dynamics of traffic scenarios and improve prediction precision, especially in complex environments. The proposed model showcases strong potential for application in real-world autonomous driving systems.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

56 extracted references · 49 canonical work pages

  1. [1]

    A survey on trajectory-prediction methods for autonomous driving,

    Y . Huang, J. Du, Z. Y ang, Z. Zhou, L. Zhang, and H. Chen, “A survey on trajectory-prediction methods for autonomous driving,” IEEE Transactions on Intelligent V ehicles, vol. 7, no. 3, pp. 652–674, 2022

    work page 2022

  2. [2]

    Mftraj: Map-free, behavior-driven trajectory prediction for autonomous driving,

    H. Liao, Z. Li, C. Wang, H. Shen, B. Wang, D. Liao, G. Li, and C. Xu, “Mftraj: Map-free, behavior-driven trajectory prediction for autonomous driving,” 2024

    work page 2024

  3. [3]

    Physics-informed trajectory prediction for au- tonomous driving under missing observation,

    H. Liao, C. Wang, Z. Li, Y . Li, B. Wang, G. Li, and C. Xu, “Physics-informed trajectory prediction for au- tonomous driving under missing observation,” Available at SSRN 4809575 , 2024

    work page 2024

  4. [4]

    GPT-4 Enhanced Multimodal Grounding for Autonomous Driving: Leveraging Cross-Modal Atten- tion with Large Language Models,

    H. Liao, H. Shen, Z. Li, C. Wang, G. Li, Y . Bie, and C. Xu, “GPT-4 Enhanced Multimodal Grounding for Autonomous Driving: Leveraging Cross-Modal Atten- tion with Large Language Models,” arXiv e-prints , p. arXiv:2312.03543, Dec. 2023

    work page arXiv 2023

  5. [5]

    A Cognitive-Driven Trajectory Prediction Model for Autonomous Driving in Mixed Autonomy Environment

    H. Liao, Z. Li, C. Wang, B. Wang, H. Kong, Y . Guan, G. Li, Z. Cui, and C. Xu, “A cognitive- driven trajectory prediction model for autonomous driv- ing in mixed autonomy environment,” arXiv preprint arXiv:2404.17520, 2024

    work page Pith review arXiv 2024

  6. [6]

    Uncertainty-aware driver trajectory prediction at urban intersections,

    X. Huang, S. G. McGill, B. C. Williams, L. Fletcher, and G. Rosman, “Uncertainty-aware driver trajectory prediction at urban intersections,” in 2019 International conference on robotics and automation (ICRA). IEEE, 2019, pp. 9718–9724

    work page 2019

  7. [7]

    Bat: Behavior-aware human-like trajectory prediction for autonomous driving,

    H. Liao, Z. Li, H. Shen, W . Zeng, D. Liao, G. Li, and C. Xu, “Bat: Behavior-aware human-like trajectory prediction for autonomous driving,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 38, no. 9, 2024, pp. 10 332–10 340

    work page 2024

  8. [8]

    Gpt-4 enhanced multimodal grounding for au- tonomous driving: Leveraging cross-modal attention with large language models,

    H. Liao, H. Shen, Z. Li, C. Wang, G. Li, Y . Bie, and C. Xu, “Gpt-4 enhanced multimodal grounding for au- tonomous driving: Leveraging cross-modal attention with large language models,” Communications in Trans- portation Research, vol. 4, p. 100116, 2024

    work page 2024

Show all 56 references
  1. [9]

    Human observation-inspired trajectory prediction for autonomous driving in mixed-autonomy traffic environments,

    H. Liao, S. Liu, Y . Li, Z. Li, C. Wang, B. Wang, Y . Guan, and C. Xu, “Human observation-inspired trajectory prediction for autonomous driving in mixed-autonomy traffic environments,” arXiv preprint arXiv:2402.04318, 2024

    2024 arXiv

  2. [10]

    A cognitive-based trajectory prediction ap- proach for autonomous driving,

    H. Liao, Y . Li, Z. Li, C. Wang, Z. Cui, S. E. Li, and C. Xu, “A cognitive-based trajectory prediction ap- proach for autonomous driving,” IEEE Transactions on Intelligent V ehicles, pp. 1–12, 2024

    2024

  3. [11]

    Diverse and admissible trajectory forecasting through multi- modal context understanding,

    S. H. Park, G. Lee, J. Seo, M. Bhat, M. Kang, J. Fran- cis, A. Jadhav, P . P . Liang, and L.-P . Morency, “Diverse and admissible trajectory forecasting through multi- modal context understanding,” in Computer Vision– ECCV 2020: 16th European Conference, Glasgow, UK, August 23...

    2020

  4. [12]

    Extended kalman filter for state estimation and trajectory predic- tion of a moving object detected by an unmanned aerial vehicle,

    C. G. Prevost, A. Desbiens, and E. Gagnon, “Extended kalman filter for state estimation and trajectory predic- tion of a moving object detected by an unmanned aerial vehicle,” in 2007 American control conference. IEEE, 2007, pp. 1805–1810

    2007

  5. [13]

    Probabilistic vehicle trajectory predic- tion over occupancy grid map via recurrent neural net- work,

    B. Kim, C. M. Kang, J. Kim, S. H. Lee, C. C. Chung, and J. W . Choi, “Probabilistic vehicle trajectory predic- tion over occupancy grid map via recurrent neural net- work,” in IEEE ITSC. IEEE, 2017, pp. 399–404

    2017

  6. [14]

    An lstm network for highway trajectory prediction,

    F. Altch´ e and A. de La Fortelle, “An lstm network for highway trajectory prediction,” in IEEE 20th ITSC . IEEE, 2017, pp. 353–359

    2017

  7. [15]

    Social lstm: Human tra- jectory prediction in crowded spaces,

    A. Alahi, K. Goel, V . Ramanathan, A. Robicquet, L. Fei-Fei, and S. Savarese, “Social lstm: Human tra- jectory prediction in crowded spaces,” in Proceedings of the IEEE CVPR , 2016, pp. 961–971

    2016

  8. [16]

    Ast-gnn: An attention-based spatio- temporal graph neural network for interaction-aware pedestrian trajectory prediction,

    H. Zhou, D. Ren, H. Xia, M. Fan, X. Y ang, and H. Huang, “Ast-gnn: An attention-based spatio- temporal graph neural network for interaction-aware pedestrian trajectory prediction,” Neurocomputing, vol. 445, pp. 298–308, 2021

    2021

  9. [17]

    Goal-gan: Multimodal trajectory prediction based on goal position estimation,

    P . Dendorfer, A. Osep, and L. Leal-Taix´ e, “Goal-gan: Multimodal trajectory prediction based on goal position estimation,” in Proceedings of the Asian Conference on Computer Vision, 2020

    2020

  10. [18]

    Desire: Distant future prediction in dynamic scenes with interacting agents,

    N. Lee, W . Choi, P . V ernaza, C. B. Choy, P . H. Torr, and M. Chandraker, “Desire: Distant future prediction in dynamic scenes with interacting agents,” in Proceed- ings of the IEEE CVPR , 2017, pp. 336–345

    2017

  11. [19]

    Hierarchical text-conditional im- age generation with clip latents,

    A. Ramesh, P . Dhariwal, A. Nichol, C. Chu, and M. Chen, “Hierarchical text-conditional im- age generation with clip latents,” arXiv preprint arXiv:2204.06125, vol. 1, no. 2, p. 3, 2022

    2022 arXiv

  12. [20]

    High-resolution image synthesis with la- tent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P . Esser, and B. Ommer, “High-resolution image synthesis with la- tent diffusion models,” in Proceedings of the IEEE/CVF CVPR, 2022, pp. 10 684–10 695

    2022

  13. [21]

    Photorealistic text- to-image diffusion models with deep language under- standing,

    C. Saharia, W . Chan, S. Saxena, L. Li, J. Whang, E. L. Denton, K. Ghasemipour, R. Gontijo Lopes, B. Karagol A yan, T. Salimanset al., “Photorealistic text- to-image diffusion models with deep language under- standing,” Advances in Neural Information Processing Systems, vol. 3...

    2022

  14. [22]

    Imagen video: High definition video generation with diffusion models,

    J. Ho, W . Chan, C. Saharia, J. Whang, R. Gao, A. Gritsenko, D. P . Kingma, B. Poole, M. Norouzi, D. J. Fleet et al. , “Imagen video: High definition video generation with diffusion models,” arXiv preprint arXiv:2210.02303, 2022

    arXiv

  15. [23]

    Diffusion prob- abilistic modeling for video generation,

    R. Y ang, P . Srivastava, and S. Mandt, “Diffusion prob- abilistic modeling for video generation,” arXiv preprint arXiv:2203.09481, 2022

    2022 arXiv

  16. [24]

    Dreamfusion: Text-to-3d using 2d diffusion,

    B. Poole, A. Jain, J. T. Barron, and B. Mildenhall, “Dreamfusion: Text-to-3d using 2d diffusion,” arXiv preprint arXiv:2209.14988, 2022

    2022 arXiv

  17. [25]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P . Abbeel, “Denoising diffusion probabilistic models,” Advances in neural information processing systems, vol. 33, pp. 6840–6851, 2020

    2020

  18. [26]

    Deep unsupervised learning using nonequilibrium thermodynamics,

    J. Sohl-Dickstein, E. Weiss, N. Maheswaranathan, and S. Ganguli, “Deep unsupervised learning using nonequilibrium thermodynamics,” in International con- ference on machine learning. PMLR, 2015, pp. 2256– 2265

    2015

  19. [27]

    Social gan: Socially acceptable trajectories with generative adversarial networks,

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

    2018

  20. [28]

    Convolutional social pool- ing for vehicle trajectory prediction,

    N. Deo and M. M. Trivedi, “Convolutional social pool- ing for vehicle trajectory prediction,” in Proceedings of the IEEE CVPR W orkshops, 2018, pp. 1468–1476

    2018

  21. [29]

    Multi-agent tensor fusion for contextual trajectory prediction,

    T. Zhao, Y . Xu, M. Monfort, W . Choi, C. Baker, Y . Zhao, Y . Wang, and Y . N. Wu, “Multi-agent tensor fusion for contextual trajectory prediction,” in Proceed- ings of the IEEE/CVF CVPR , 2019, pp. 12 126–12 134

    2019

  22. [30]

    Non-local social pooling for vehicle tra- jectory prediction,

    K. Messaoud, I. Y ahiaoui, A. V erroust-Blondet, and F. Nashashibi, “Non-local social pooling for vehicle tra- jectory prediction,” in IEEE IV. IEEE, 2019, pp. 975– 980

    2019

  23. [31]

    Multi-modal trajectory pre- diction of surrounding vehicles with maneuver based lstms,

    N. Deo and M. M. Trivedi, “Multi-modal trajectory pre- diction of surrounding vehicles with maneuver based lstms,” in IEEE IV. IEEE, 2018, pp. 1179–1184

    2018

  24. [32]

    Interaction-aware motion prediction for au- tonomous driving: A multiple model kalman filtering scheme,

    V . Lefkopoulos, M. Menner, A. Domahidi, and M. N. Zeilinger, “Interaction-aware motion prediction for au- tonomous driving: A multiple model kalman filtering scheme,” IEEE Robotics and Automation Letters, vol. 6, no. 1, pp. 80–87, 2020

    2020

  25. [33]

    Imitating driver behavior with generative adversar- ial networks,

    A. Kuefler, J. Morton, T. Wheeler, and M. Kochender- fer, “Imitating driver behavior with generative adversar- ial networks,” in IEEE IV. IEEE, 2017, pp. 204–211

    2017

  26. [34]

    Multiple futures prediction,

    C. Tang and R. R. Salakhutdinov, “Multiple futures prediction,” Advances in neural information processing systems, vol. 32, 2019

    2019

  27. [35]

    Dual transformer based prediction for lane change in- tentions and trajectories in mixed traffic environment,

    K. Gao, X. Li, B. Chen, L. Hu, J. Liu, R. Du, and Y . Li, “Dual transformer based prediction for lane change in- tentions and trajectories in mixed traffic environment,” IEEE Transactions on Intelligent Transportation Sys- tems, 2023

    2023

  28. [36]

    Neighbourhood context embeddings in deep inverse reinforcement learning for predicting pedestrian mo- tion over long time horizons,

    T. Fernando, S. Denman, S. Sridharan, and C. Fookes, “Neighbourhood context embeddings in deep inverse reinforcement learning for predicting pedestrian mo- tion over long time horizons,” in Proceedings of the IEEE/CVF ICCV W orkshops, 2019, pp. 0–0

    2019

  29. [37]

    Wsip: Wave superposition inspired pooling for dynamic interactions-aware trajectory prediction,

    R. Wang, S. Wang, H. Y an, and X. Wang, “Wsip: Wave superposition inspired pooling for dynamic interactions-aware trajectory prediction,” in Proceed- ings of the AAAI Conference on Artificial Intelligence , vol. 37, no. 4, 2023, pp. 4685–4692

    2023

  30. [38]

    Congestion-aware multi-agent trajectory prediction for collision avoidance,

    X. Xie, C. Zhang, Y . Zhu, Y . N. Wu, and S.-C. Zhu, “Congestion-aware multi-agent trajectory prediction for collision avoidance,” in 2021 IEEE International Con- ference on Robotics and Automation (ICRA) . IEEE, 2021, pp. 13 693–13 700

    2021

  31. [39]

    Attention based vehicle trajectory pre- diction,

    K. Messaoud, I. Y ahiaoui, A. V erroust-Blondet, and F. Nashashibi, “Attention based vehicle trajectory pre- diction,” IEEE Transactions on Intelligent V ehicles , vol. 6, no. 1, pp. 175–185, 2021

    2021

  32. [40]

    Hier- archical motion encoder-decoder network for trajectory forecasting,

    Q. Xue, S. Li, X. Li, J. Zhao, and W . Zhang, “Hier- archical motion encoder-decoder network for trajectory forecasting,” arXiv preprint arXiv:2111.13324 , 2021

    2021 arXiv

  33. [41]

    Multi-vehicle collaborative learning for trajectory pre - diction with spatio-temporal tensor fusion,

    Y . Wang, S. Zhao, R. Zhang, X. Cheng, and L. Y ang, “Multi-vehicle collaborative learning for trajectory pre - diction with spatio-temporal tensor fusion,” IEEE Transactions on Intelligent Transportation Systems , vol. 23, no. 1, pp. 236–248, 2022

    2022

  34. [42]

    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. Y ang, “Intention-aware vehicle trajectory prediction based on spatial-temporal dynamic attention network for internet of vehicles,” IEEE Transactions on Intelligent Trans- portation Systems , vol. 23, no. 10, pp. 19 471–19 483, 2022

    2022

  35. [43]

    Environment-attention network for vehicle trajectory prediction,

    Y . Cai, Z. Wang, H. Wang, L. Chen, Y . Li, M. A. Sotelo, and Z. Li, “Environment-attention network for vehicle trajectory prediction,” IEEE Transactions on V ehicular T echnology, vol. 70, no. 11, pp. 11 216–11 227, 2021

    2021

  36. [44]

    Empirical evalu- ation of rectified activations in convolutional network,

    B. Xu, N. Wang, T. Chen, and M. Li, “Empirical evalu- ation of rectified activations in convolutional network,” arXiv preprint arXiv:1505.00853 , 2015

    2015 arXiv

  37. [45]

    An introductory survey on attention mecha- nisms in nlp problems,

    D. Hu, “An introductory survey on attention mecha- nisms in nlp problems,” in Intelligent Systems and Ap- plications: Proceedings of the 2019 Intelligent Systems Conference (IntelliSys) V olume 2. Springer, 2020, pp. 432–448

    2019

  38. [46]

    Evolveg- raph: Multi-agent trajectory prediction with dynamic relational reasoning,

    J. Li, F. Y ang, M. Tomizuka, and C. Choi, “Evolveg- raph: Multi-agent trajectory prediction with dynamic relational reasoning,” Advances in neural information processing systems, vol. 33, pp. 19 783–19 794, 2020

    2020

  39. [47]

    The highd dataset: A drone dataset of naturalistic ve- hicle 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 ve- hicle trajectories on german highways for validation of highly automated driving systems,” in 21st ITSC . IEEE, 2018, pp. 2118–2125

    2018

  40. [48]

    Multiple knowledge representation of artificial intelligence,

    Y . Pan, “Multiple knowledge representation of artificial intelligence,” Engineering, vol. 6, no. 3, pp. 216–217, 2020

    2020

  41. [49]

    Multi- head attention based probabilistic vehicle trajectory pre- diction,

    H. Kim, D. Kim, G. Kim, J. Cho, and K. Huh, “Multi- head attention based probabilistic vehicle trajectory pre- diction,” in 2020 IEEE Intelligent V ehicles Symposium (IV). IEEE, 2020, pp. 1720–1725

    2020

  42. [50]

    Multi-vehicle collaborative learning for trajectory pre - diction with spatio-temporal tensor fusion,

    Y . Wang, S. Zhao, R. Zhang, X. Cheng, and L. Y ang, “Multi-vehicle collaborative learning for trajectory pre - diction with spatio-temporal tensor fusion,” IEEE Transactions on Intelligent Transportation Systems , vol. 23, no. 1, pp. 236–248, 2020

    2020

  43. [51]

    Root mean square error (rmse) or mean absolute error (mae),

    T. Chai and R. R. Draxler, “Root mean square error (rmse) or mean absolute error (mae),” Geoscientific model development discussions, vol. 7, no. 1, pp. 1525– 1534, 2014

    2014

  44. [52]

    Ss-lstm: A hierarchical lstm model for pedestrian trajectory predic- tion,

    H. Xue, D. Q. Huynh, and M. Reynolds, “Ss-lstm: A hierarchical lstm model for pedestrian trajectory predic- tion,” in 2018 IEEE Winter Conference on Applications of Computer Vision (WACV) . IEEE, 2018, pp. 1186– 1194

    2018

  45. [53]

    Encoding crowd interac- tion with deep neural network for pedestrian trajectory prediction,

    Y . Xu, Z. Piao, and S. Gao, “Encoding crowd interac- tion with deep neural network for pedestrian trajectory prediction,” in Proceedings of the IEEE CVPR , 2018, pp. 5275–5284

    2018

  46. [54]

    Dacr-amtp: adaptive multi-modal vehicle trajectory prediction for dynamic drivable areas based on collision risk,

    P . Cong, Y . Xiao, X. Wan, M. Deng, J. Li, and X. Zhang, “Dacr-amtp: adaptive multi-modal vehicle trajectory prediction for dynamic drivable areas based on collision risk,” IEEE Transactions on Intelligent V ehicles, 2023

    2023

  47. [55]

    Trajectory prediction network of autonomous vehi- cles with fusion of historical interactive features,

    Z. Zuo, X. Wang, S. Guo, Z. Liu, Z. Li, and Y . Wang, “Trajectory prediction network of autonomous vehi- cles with fusion of historical interactive features,” IEEE Transactions on Intelligent V ehicles, 2023

    2023

  48. [56]

    Autonomous driving roadway feature in- terpretation using integrated semantic analysis and do- main adaptation,

    S. Xi, Z. Liu, Z. Wang, Q. Zhang, H. Ding, C. C. Kang, and Z. Chen, “Autonomous driving roadway feature in- terpretation using integrated semantic analysis and do- main adaptation,” IEEE Access, 2024. This figure "cdffusion1.png" is available in "png" format from: http://arxiv...

    2024 arXiv

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.