Pith. sign in

REVIEW 3 major objections 4 minor 33 references

EPN: An Ego Vehicle Planning-Informed Network for Target Trajectory Prediction

T0 review · 3 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash

Pith's one-line read EPN predicts target trajectories by conditioning on the ego vehicle's planned future path and on corrected target endpoints, reporting error reductions of about 35% on NGSIM and 64% on HighD over the strongest baselines tested.

desk verdict Solid incremental architecture, but the paper never says where the ego planned trajectory comes from on datasets that have no planner, so the headline gains are uninterpretable. read the letter →

arxiv 2412.14442 v2 pith:R24TJU4P submitted 2024-12-19 cs.RO

classification cs.RO
keywords trajectorypredictionegovehicleplanningmultimodalendpointconditionalvariationalautoencoderNGSIMHighDautonomousdriving
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 argues that trajectory prediction for a target vehicle improves when the predictor is given the ego vehicle's own planned future trajectory as an extra input, alongside a module that first predicts where the target will end up and then corrects that endpoint. It claims this combination, called EPN, produces average reductions of 34.9% in RMSE, 30.7% in ADE, and 30.4% in FDE on NGSIM, and 64.6%, 64.5%, and 64.3% on HighD, compared with the strongest baselines in the experiments. If that is right, feeding planned ego motion and endpoint refinement is a practical way to cut prediction error on highway datasets without relying on high-definition maps.

What carries the argument

The load-bearing mechanism is the pairing of two inputs: the ego planned trajectory $P$ (Eq. 3), fed through its own LSTM and social-pooling branch, and a target endpoint prediction module that uses a conditional variational autoencoder (CVAE) to sample endpoints, a correction decoder that adds an offset $d_{\text{offset}}$ to the sampled endpoint, and an LSTM decoder that conditions full trajectories on the corrected endpoint. The correction mechanism is what separates EPN from plain endpoint-conditioned prediction; the ablation attributes a 14.1%, 18.4%, and 21.5% reduction in RMSE, ADE, and FDE to adding the endpoint module and correction on top of the planning-informed baseline.

What would settle it

Take a test split, replace P with a planner output computed only from history (for example, constant-velocity extrapolation or a learned policy), retrain and reevaluate EPN against PiP on the same metrics; if the large error reductions disappear, the reported gains were carried by leaked future information rather than by the architecture.

Watch

Extended reading notes

Core claim

The central claim is that multimodal target trajectory prediction errors drop substantially when the predictor conditions on the ego vehicle's planned future trajectory and on corrected target endpoints rather than on discrete driving intentions. The model encodes historical states of the ego, target, and neighboring vehicles through LSTM encoders, pools them with a convolutional social network, and also encodes the ego plan; a conditional variational autoencoder samples candidate endpoints, a correction head predicts an offset between the sampled and true endpoint, and an LSTM decoder turns each corrected endpoint into a full trajectory. The authors report that EPN beats the prior planning-informed model PiP and the intention-based baselines by large margins, with the largest gains on the drone-recorded HighD dataset.

Load-bearing premise

The experiments treat the ego vehicle's planned future trajectory P as known input on datasets that do not contain planner outputs, and the paper does not say how P was produced; if P is actually the ground-truth future ego trajectory, the comparison hands EPN future information that the baselines lack.

Editorial extensions

If this is right

  • If P is genuinely available at test time, planning-informed prediction can be applied on mapless highway datasets and outperform history-only and intention-based models.
  • Endpoint correction reduces error more than simply adding more trajectory modalities, according to the ablation results.
  • On HighD, where positional data is more accurate, the plan encoding and endpoint module cut errors by roughly two-thirds, suggesting better sensor data amplifies the benefit.
  • The model produces multimodal predictions by sampling the latent space multiple times, with $k=6$ endpoints matching the six intention classes used by the baselines.

Reading between the lines

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

  • A reader should treat the headline numbers as conditional on the ego planned trajectory P being produced honestly from history, since neither NGSIM nor HighD contains planner outputs and the paper does not describe how P was generated.
  • One direct test of the claim is to replace P with a constant-velocity or kinematic extrapolation from history alone and see whether the reported reductions persist; if they vanish, the gains were carried by future information rather than by the architecture.
  • The endpoint-correction idea could be attached to other endpoint-driven predictors beyond CVAE, since the correction decoder only needs an initial endpoint estimate and an environmental feature vector.
  • Deploying EPN in a closed-loop simulator, where the ego plan is genuinely known before execution, would clarify whether the planning input helps in real systems rather than only in offline evaluation.
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

3 major / 4 minor

Summary. The paper proposes EPN, a multimodal trajectory prediction network that augments a social-pooling LSTM architecture with the ego vehicle's future planned trajectory P (Eq. (3)) and with a CVAE-based endpoint prediction module that corrects sampled endpoints before decoding complete trajectories. Experiments on NGSIM and HighD report large reductions in RMSE, ADE, and FDE relative to S-LSTM, CS-LSTM, S-GAN, WSiP, PiP, and S-TF, and ablations on NGSIM attribute part of the gain to speed/acceleration inputs, to the endpoint prediction module, and to the endpoint correction mechanism. Code is promised only after acceptance.

Significance. If the input P is produced causally, the core idea of conditioning target prediction on an ego plan is relevant and the endpoint-correction design is a reasonable engineering contribution. The ablations are a strength: they isolate the contributions of speed, acceleration, endpoint prediction, and endpoint correction, and the architecture description is coherent. However, the paper's headline significance rests almost entirely on the reported performance gap, and that gap depends on a key input whose source is not specified for either dataset. The manuscript therefore does not currently provide an interpretable comparison of EPN against the baselines, and its reproducibility is limited by the absence of code and by the unresolved construction of P.

major comments (3)
  1. [Section V-A, Eq. (3)] The paper never states how the ego planned trajectory P is generated for NGSIM and HighD. Neither dataset contains planner outputs, and Section V-A describes only downsampling, segmentation, and a 7:1:2 split. If P is taken from the recorded future positions of the vehicle treated as the ego vehicle, then EPN receives information about the future during inference while the comparison methods do not. Because Eq. (13) encodes P through an LSTM and Eq. (16) passes it through convolutional pooling into encsocial, the network could exploit the strong correlation between the ego's future motion and the target's future motion, which would make the reported gains an artifact of information leakage. The authors must specify the exact causal procedure that produces P using only information available at time t, and must rerun the comparisons under that procedure; without this, the central empirical claim is uninterpretable.
  2. [Section V-C, Tables I-III] The baseline numbers are presented without implementation details or sources, so the comparison cannot be reproduced. In particular, S-TF is reported with RMSE values (0.99 at 1s on NGSIM, 0.75 at 1s on HighD) that are much worse than the other learned baselines, and the claimed reductions of 34.9% and 64.6% are computed against this method. The authors should provide the exact evaluation protocol, the hyperparameters, and either code or a precise citation of where each tabulated value comes from. Otherwise the headline reductions and the conclusions drawn from them are not verifiable.
  3. [Section V-C, Table III] The multimodal evaluation selects for each method the trajectory closest to the true future trajectory before computing ADE and FDE. This oracle selection is common in the literature, but the number of candidate trajectories must be matched across methods. The paper states k=6 endpoints, yet the CVAE sampling procedure in Section IV-B can generate any number of endpoint samples via multiple draws from the latent space. The authors should state the total number of output trajectories used for EPN and for each baseline, so that the comparison is not biased by a larger candidate set.
minor comments (4)
  1. [Throughout, Tables I-II] "RSME" is a typo and should be "RMSE" throughout the tables and text.
  2. [Section IV-A, Eq. (13)] Equation (13) uses the variables x_i^t, y_i^t for the ego planned trajectory, but P in Eq. (3) is defined as p_t; the notation should be made consistent so that the ego plan coordinates are clearly distinguished from those of other vehicles.
  3. [Section III-A] The symbols Atar and Aner are used for the region of interest, but later equations use Anei for the set of neighboring vehicles; these notations should be unified.
  4. [Section V-A] The paper refers to a 7:1:2 split by scenario but does not state whether the split is performed at the track/segment level or the trajectory level; this should be clarified for reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: EPN's prediction is a supervised empirical claim; the unspecified source of ego planned trajectory P is a potential information-leakage threat, not a circular derivation.

full rationale

EPN is a supervised trajectory prediction model trained with Lpred and Lcvae on held-out NGSIM/HighD data, so its central performance claim is empirical and is not derived from its inputs by construction. The endpoint module uses true endpoints only during training (Eqs. 19-21), and the closest-to-truth selection for multimodal evaluation is explicitly disclosed in Section V-C as 'selecting the trajectory closest to the true future trajectory for evaluation'; this is a standard, though optimistic, evaluation choice rather than a fitted parameter renamed as a prediction. I found no self-citation chain, no imported uniqueness theorem, and no ansatz smuggled in via citation. The main substantive concern, flagged by the skeptic, is that the planned ego trajectory P in Eq. (3) is never sourced for NGSIM/HighD; if P is taken from the recorded future ego trajectory, EPN would be given future information unavailable to the baselines. That is a validity and fairness threat, not circularity: the model still computes target predictions from its inputs through learned network functions, and the reported reductions are not equivalent to an input by definition. The missing description of P should be corrected before the comparison can be interpreted, but it does not make the derivation circular, so the circularity score is 0.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The central empirical claim rests on an unstated input-generation protocol for P, a fixed number of samples, and an oracle metric. These are not derived, so they must be counted as assumptions. The learned weights are fitted to training data, which is normal supervised learning, but the key hyperparameters and the plan source are hand choices that materially affect the headline numbers.

free parameters (4)
  • CVAE sampling standard deviation sigma_T = 1.3
    Hand-chosen following PECnet's truncation trick; controls diversity of sampled endpoints and affects best-of-k metrics.
  • Number of endpoints K = 6
    Set to match the six driving-intention classes of baselines; larger K mechanically improves minADE and minFDE.
  • History and prediction horizon = Th=15, Tf=25 at 5 Hz
    Downsampled from original 10 Hz and 25 Hz datasets; no sensitivity analysis for this choice.
  • Network hyperparameters = LSTM 64/128, lr 0.001, 15 epochs, batch 64
    Reported without ablations or multiple seeds; the empirical claims may depend on these choices.
assumptions (4)
  • domain assumption The ego planned trajectory P is available at inference and does not encode ground-truth future information.
    Section III-A defines P as a model input, but Section V-A never states how P is constructed for NGSIM and HighD, which contain no planner outputs.
  • domain assumption Historical position, speed, and acceleration of surrounding vehicles are sufficient to predict the target vehicle's future.
    Section IV-A encodes only these quantities; no map, lane, or traffic-rule information is used.
  • ad hoc to paper Selecting the closest predicted trajectory to the ground truth is an acceptable evaluation of predictive skill.
    Section V-C evaluates minADE/minFDE style metrics; the selected mode cannot be identified at inference time, inflating reported accuracy.
  • domain assumption The six CVAE samples adequately cover the target's multimodal future distribution.
    Section IV-B fixes K=6 with sigma_T=1.3; no coverage or miss-rate analysis is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of EPN: An Ego Vehicle Planning-Informed Network for Target Trajectory Prediction." pith.science (2026). https://pith.science/paper/R24TJU4P

@misc{pith2026241214442,
  author       = {Pith},
  title        = {Pith review of: EPN: An Ego Vehicle Planning-Informed Network for Target Trajectory Prediction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/R24TJU4P}},
  note         = {Machine review of arXiv:2412.14442}
}
read the original abstract

Trajectory prediction plays a crucial role in improving the safety of autonomous vehicles. However, due to the highly dynamic and multimodal nature of the task, accurately predicting the future trajectory of a target vehicle remains a significant challenge. To address this challenge, we propose an Ego vehicle Planning-informed Network (EPN) for multimodal trajectory prediction. In real-world driving, the future trajectory of a vehicle is influenced not only by its own historical trajectory, but also by the behavior of other vehicles. So, we incorporate the future planned trajectory of the ego vehicle as an additional input to simulate the mutual influence between vehicles. Furthermore, to tackle the challenges of intention ambiguity and large prediction errors often encountered in methods based on driving intentions, we propose an endpoint prediction module for the target vehicle. This module predicts the target vehicle endpoints, refines them using a correction mechanism, and generates a multimodal predicted trajectory. Experimental results demonstrate that EPN achieves an average reduction of 34.9%, 30.7%, and 30.4% in RMSE, ADE, and FDE on the NGSIM dataset, and an average reduction of 64.6%, 64.5%, and 64.3% in RMSE, ADE, and FDE on the HighD dataset. The code will be open sourced after the letter is accepted.

Figures

Figures reproduced from arXiv: 2412.14442 by the authors.

Figure 1
Figure 1. Vehicle classification and region of interest division in a driving [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Overall architecture of EPN model. encsocial, which encapsulates the spatial feature information of the target vehicle. The formulas are as follows: encnei = ReLu(Conv2(MP1(ReLu(Conv1(tnei))))) (15) encplan = ReLu(Conv2(MP1(ReLu(Conv1(tplan))))) (16) encsocial = MP2(concat(encnei, encplan)) (17) Finally, we combine the dynamic feature vector of the target vehicle with the social feature vector to create the environm… view at source ↗
Figure 3
Figure 3. Visualization results in left-turn, straight-ahead and right-turn [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references · 33 canonical work pages

  1. [1]

    A survey on trajectory-prediction methods for autonomous driving,

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

  2. [2]

    Vehicle dynamics and external disturbance estimation for vehicle path prediction,

    C.-F. Lin, A. Ulsoy, and D. LeBlanc, “Vehicle dynamics and external disturbance estimation for vehicle path prediction,”IEEE Trans. Auton. Mental Develop. , vol. 8, no. 3, pp. 508–518, 2000

  3. [3]

    Situation assessment of an autonomous emergency brake for arbitrary vehicle-to-vehicle collision scenarios,

    N. Kaempchen, B. Schiele, and K. Dietmayer, “Situation assessment of an autonomous emergency brake for arbitrary vehicle-to-vehicle collision scenarios,” IEEE Trans. Intell. Transport. Syst., vol. 10, no. 4, pp. 678–687, 2009

  4. [4]

    Comparison and evaluation of advanced motion models for vehicle tracking,

    R. Schubert, E. Richter, and G. Wanielik, “Comparison and evaluation of advanced motion models for vehicle tracking,” in Int. Conf. Inf. Fusion, 2008, pp. 1–6

  5. [5]

    Cooperative path pre- diction in vehicular environments,

    P. Lytrivis, G. Thomaidis, and A. Amditis, “Cooperative path pre- diction in vehicular environments,” in Int. IEEE Conf. Intell. Transp. Syst., 2008, pp. 803–808

  6. [6]

    Switched kalman filter- interacting multiple model algorithm based on optimal autoregressive model for manoeuvring target tracking,

    B. Jin, B. Jiu, T. Su, H. Liu, and G. Liu, “Switched kalman filter- interacting multiple model algorithm based on optimal autoregressive model for manoeuvring target tracking,” IET Radar , Sonar & Navig. , vol. 9, no. 2, pp. 199–209, 2015

  7. [7]

    Object tracking in urban intersections based on active use of a priori knowledge: Active interacting multi model filter,

    H. Dyckmanns, R. Matthaei, M. Maurer, B. Lichte, J. Effertz, and D. St ¨uker, “Object tracking in urban intersections based on active use of a priori knowledge: Active interacting multi model filter,” in IEEE Intell. V eh. Symp., 2011, pp. 625–630

  8. [8]

    Trajectory planning and safety assessment of autonomous vehicles based on motion prediction and model predictive control,

    Y . Wang, Z. Liu, Z. Zuo, Z. Li, L. Wang, and X. Luo, “Trajectory planning and safety assessment of autonomous vehicles based on motion prediction and model predictive control,” IEEE Trans. V eh. Technol., vol. 68, no. 9, pp. 8546–8556, 2019

Show all 33 references
  1. [9]

    Modeling multi-vehicle interaction scenarios using gaussian random field,

    Y . Guo, V . V . Kalidindi, M. Arief, W. Wang, J. Zhu, H. Peng, and D. Zhao, “Modeling multi-vehicle interaction scenarios using gaussian random field,” in IEEE Intell. Transp. Syst. Conf. , 2019, pp. 3974– 3980

  2. [10]

    Threat assessment design for driver assistance system at intersections,

    G. S. Aoude, B. D. Luders, K. K. H. Lee, D. S. Levine, and J. P. How, “Threat assessment design for driver assistance system at intersections,” in 13th Int. IEEE Conf. Intell. Transp. Syst. , 2010, pp. 1855–1862

  3. [11]

    A self-adaptive parameter selection trajectory prediction approach via hidden markov models,

    S. Qiao, D. Shen, X. Wang, N. Han, and W. Zhu, “A self-adaptive parameter selection trajectory prediction approach via hidden markov models,” IEEE Trans. Intell. Transport. Syst. , vol. 16, no. 1, pp. 284– 296, 2015

  4. [12]

    A dynamic bayesian network for vehicle maneuver prediction in highway driving scenarios: Framework and verification,

    J. Li, B. Dai, X. Li, X. Xu, and D. Liu, “A dynamic bayesian network for vehicle maneuver prediction in highway driving scenarios: Framework and verification,” Electronics, vol. 8, no. 1, 2019

  5. [13]

    Context-aware timewise vaes for real-time vehicle trajectory prediction,

    P. Xu, J.-B. Hayet, and I. Karamouzas, “Context-aware timewise vaes for real-time vehicle trajectory prediction,” IEEE Robot. Automat. Lett., vol. 8, no. 9, pp. 5440–5447, 2023

  6. [14]

    Convolutional social pooling for vehicle trajectory prediction,

    N. Deo and M. M. Trivedi, “Convolutional social pooling for vehicle trajectory prediction,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. Workshops, June 2018

  7. [15]

    So- cial gan: Socially acceptable trajectories with generative adversarial networks,

    A. Gupta, J. Johnson, L. Fei-Fei, S. Savarese, and A. Alahi, “So- cial gan: Socially acceptable trajectories with generative adversarial networks,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. , June 2018

  8. [16]

    Vehicle trajectory prediction using gen- erative adversarial network with temporal logic syntax tree features,

    X. Li, G. Rosman, I. Gilitschenski, C.-I. Vasile, J. A. DeCastro, S. Karaman, and D. Rus, “Vehicle trajectory prediction using gen- erative adversarial network with temporal logic syntax tree features,” IEEE Robot. Automat. Lett. , vol. 6, no. 2, pp. 3459–3466, 2021

  9. [17]

    Intention-aware long horizon trajectory prediction of surrounding vehicles using dual lstm networks,

    L. Xin, P. Wang, C.-Y . Chan, J. Chen, S. E. Li, and B. Cheng, “Intention-aware long horizon trajectory prediction of surrounding vehicles using dual lstm networks,” in 2018 21st Int. Conf. Intell. Transp. Syst., 2018, pp. 1441–1446

  10. [18]

    Improving multi-agent trajectory prediction using traffic states on interactive driving scenarios,

    C. Vishnu, V . Abhinav, D. Roy, C. K. Mohan, and C. S. Babu, “Improving multi-agent trajectory prediction using traffic states on interactive driving scenarios,” IEEE Robot. Automat. Lett. , vol. 8, no. 5, pp. 2708–2715, 2023

  11. [19]

    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,” in Proc. Eur . Conf. Comput. Vis. Springer, 2020, pp. 598–614

  12. [20]

    Vehicle trajectory prediction method coupled with ego vehicle motion trend under dual attention mechanism,

    H. Guo, Q. Meng, D. Cao, H. Chen, J. Liu, and B. Shang, “Vehicle trajectory prediction method coupled with ego vehicle motion trend under dual attention mechanism,” IEEE Trans. Instrum. and Meas. , vol. 71, pp. 1–16, 2022

  13. [21]

    Ego-planning-guided multi-graph convolutional network for heterogeneous agent trajectory prediction,

    Z. Sheng, Z. Huang, and S. Chen, “Ego-planning-guided multi-graph convolutional network for heterogeneous agent trajectory prediction,” Computer-Aided Civil and Infrastructure Engineering , vol. 39, no. 22, pp. 3357–3374, 2024

  14. [22]

    Macformer: Map-agent coupled transformer for real-time and robust trajectory prediction,

    C. Feng, H. Zhou, H. Lin, Z. Zhang, Z. Xu, C. Zhang, B. Zhou, and S. Shen, “Macformer: Map-agent coupled transformer for real-time and robust trajectory prediction,” IEEE Robot. Automat. Lett. , vol. 8, no. 10, pp. 6795–6802, 2023

  15. [23]

    Multimodal trajectory prediction based on sparse weight sharing,

    Z. Liu, W. Li, S. Lin, C. Li, X. Fan, and X. Zhao, “Multimodal trajectory prediction based on sparse weight sharing,” China Journal of Highway and Transport , vol. 36, no. 9, pp. 244–256, 2023

  16. [24]

    Crat-pred: Vehicle trajectory prediction with crystal graph convolutional neural networks and multi-head self-attention,

    J. Schmidt, J. Jordan, F. Gritschneder, and K. Dietmayer, “Crat-pred: Vehicle trajectory prediction with crystal graph convolutional neural networks and multi-head self-attention,” in Int. Conf. Robot. Automat. , 2022, pp. 7799–7805

  17. [25]

    Densetnt: End-to-end trajectory predic- tion from dense goal sets,

    J. Gu, C. Sun, and H. Zhao, “Densetnt: End-to-end trajectory predic- tion from dense goal sets,” in Proc. IEEE/CVF Int. Conf. Comput. Vis., October 2021, pp. 15 303–15 312

  18. [26]

    Adapt: Efficient multi-agent trajectory prediction with adaptation,

    G. Aydemir, A. K. Akan, and F. G ¨uney, “Adapt: Efficient multi-agent trajectory prediction with adaptation,” in Proc. IEEE/CVF Int. Conf. Comput. Vis., October 2023, pp. 8295–8305

  19. [27]

    Destine: Dynamic goal queries with temporal transductive alignment for trajectory pre- diction,

    R. Karim, S. M. A. Shabestary, and A. Rasouli, “Destine: Dynamic goal queries with temporal transductive alignment for trajectory pre- diction,” in IEEE Int. Conf. Robot. Automat. , 2024, pp. 2230–2237

  20. [28]

    It is not the journey but the destination: Endpoint conditioned trajectory prediction,

    K. Mangalam, H. Girase, S. Agarwal, K.-H. Lee, E. Adeli, J. Malik, and A. Gaidon, “It is not the journey but the destination: Endpoint conditioned trajectory prediction,” in Proc. Eur . Conf. Comput. Vis. Springer, 2020, pp. 759–776

  21. [29]

    Us highway 101 dataset. federal highway administration,

    J. Colyar and J. Halkias, “Us highway 101 dataset. federal highway administration,” FHW A-HRT-07-030, Tech. Rep., 2007

  22. [30]

    Us highway i-80 dataset. federal highway administration,

    J. Halkias and J. Colyar, “Us highway i-80 dataset. federal highway administration,” FHW A-HRT-06-137., Tech. Rep., 2006

  23. [31]

    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 Int. Conf. Intell. Transp. Syst. , 2018, pp. 2118–2125

  24. [32]

    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 Proc. IEEE Conf. Comput. Vis. Pattern Recognit. , June 2016

  25. [33]

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

    R. Wang, S. Wang, H. Yan, and X. Wang, “Wsip: Wave superposition inspired pooling for dynamic interactions-aware trajectory prediction,” in Proc. AAAI Artif. Intell. , vol. 37, no. 4, 2023, pp. 4685–4692

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.