Pith. sign in

REVIEW 4 major objections 6 minor 36 references

A Multi-Loss Strategy for Vehicle Trajectory Prediction: Combining Off-Road, Diversity, and Directional Consistency Losses

T0 review · 4 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 Three auxiliary losses applied to every predicted trajectory reduce off-road errors by roughly half while preserving prediction accuracy.

desk verdict Useful auxiliary losses with a real confound: the +All gains are not cleanly attributable, and the headline metrics are the same formulas as the training losses. read the letter →

arxiv 2411.19747 v1 pith:W73RMALZ submitted 2024-11-29 cs.CV cs.AIcs.LGcs.MAcs.RO

Ahmad Rahimi , Alexandre Alahi This is my paper · ORCID
classification cs.CVcs.AIcs.LGcs.MAcs.RO
keywords trajectorypredictionautonomousdrivingauxiliarylossoff-roaddirectionconsistencydiversityscenecompliancemultimodal
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

Vehicle trajectory predictors are typically trained with a "winner takes all" objective that updates only the one trajectory closest to the ground truth, leaving the other modes unrefined and often producing predictions that leave the road, oppose traffic flow, or miss plausible maneuvers. This paper proposes three differentiable auxiliary losses—Offroad Loss, Direction Consistency Error, and Diversity Loss—that supervise every prediction mode, and shows on the nuScenes and Argoverse 2 datasets with Wayformer and AutoBots baselines that they cut the offroad metric by roughly half on original scenes and by about 37% on attacked scenes. The work also demonstrates that the losses can double as evaluation metrics for realism and diversity, and that combining all three yields a balanced improvement over all quality metrics with only minor changes in standard accuracy measures. A sympathetic reader would take the central claim to be that road-rule compliance and mode diversity are trainable properties separable from accuracy, and that any existing predictor can gain them without architectural changes.

What carries the argument

The central mechanism is a set of three fully differentiable auxiliary losses that encode geometric road knowledge and are applied uniformly across all prediction modes. Offroad Loss computes a signed distance to the union of drivable-area polygons and penalizes points beyond a margin, using ray casting for point-in-polygon tests and GPU-accelerated distance computations. Direction Consistency Error measures, for each predicted point, the minimum combined distance-and-yaw gap to any centerline point, allowing flexible matching in intersections where several centerlines with different directions coexist. Mode Diversity filters out off-road trajectories and then sums pairwise distances among the remaining feasible modes, spreading predictions without pushing them off-road. The losses are used as auxiliary terms in the total loss with a hyperparameter $\alpha$ balancing their weight, and they also serve as evaluation metrics.

What would settle it

Take a model trained with these losses and evaluate it on a scene set whose drivable-area polygons come from a different, independently produced map source than the one used in training; if the off-road metric is not markedly better than a baseline trained without the losses, then the improvement reflects overfitting to the training map annotations rather than generalizable road compliance.

Watch

Extended reading notes

Core claim

The authors claim that augmenting the standard minADE objective with three auxiliary loss terms, applied to all $M$ predicted trajectories rather than only the best one, improves scene compliance and diversity across the board. Concretely, Offroad Loss uses a signed distance function to the drivable-area polygons, penalizing points outside a margin and pulling trajectories back onto the road; Direction Consistency Error matches each predicted point to the most suitable centerline by position and heading with margins, resolving the ambiguity of intersections and multi-lane roads; Diversity Loss sums pairwise distances among the feasible (on-road) modes, spreading the set to cover distinct plausible maneuvers. Across nuScenes and Argoverse 2, these losses reduce the offroad metric by roughly half in the best cases, improve directional alignment, and increase diversity, with small increases in minADE and minFDE that the authors frame as an acceptable safety-accuracy trade-off. When all three losses are combined, the models improve on every assessed quality metric relative to the baseline, and trained models also show lower offroad errors on scenes with synthetically introduced turns.

Load-bearing premise

The evaluation metrics that measure off-road, direction, and diversity are computed with the same signed-distance and centerline-matching formulas as the proposed training losses, so the reported safety improvements assume these metrics are faithful proxies for real driving compliance and that the drivable-area and centerline annotations are accurate.

Editorial extensions

If this is right

  • Any existing trajectory prediction model can be improved in road compliance and mode diversity by adding these losses, without changing the network architecture.
  • The three losses provide a standardized, differentiable set of evaluation metrics, allowing future work to measure scene compliance independently of prediction accuracy.
  • Supervising all modes counters the winner-takes-all blind spot, so less frequent but plausible maneuvers such as continuing straight at an intersection are retained in the prediction set.
  • Models trained with the losses remain more road-compliant when the scene is perturbed by naturalistic turns, indicating improved robustness to out-of-distribution inputs.
  • The small accuracy cost, such as Wayformer's minADE rising from 1.08 to 1.13 on nuScenes with all losses, is presented as a deliberate trade-off favoring safety-critical road compliance.

Reading between the lines

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

  • Because the evaluation metrics are computed with the same signed-distance and centerline formulas as the training losses, at least part of the reported improvement may reflect direct optimization of the metric itself; an independent road-compliance test with different map sources or human judgment would reveal how much of the gain is genuine behavioral change.
  • The losses could be extended to collision avoidance and kinematic feasibility, as the authors note, and the hand-tuned $\alpha$ weights could be replaced by an adaptive weighting scheme based on the model's training state.
  • The diversity loss's feasibility filter could be reused in planning cost functions to discourage redundant trajectory options during downstream decision-making, not just during prediction training.
  • The vectorized GPU implementation suggests these losses are inexpensive enough to become a standard regularization in large-scale trajectory-prediction training, potentially shifting evaluation practice toward compliance-aware metrics.
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. This paper proposes three auxiliary loss functions for multimodal vehicle trajectory prediction: Offroad Loss (Eq. 2), Direction Consistency Error (Eq. 3), and Mode Diversity Loss (Eq. 4). The losses are applied to all predicted modes rather than only the best mode, and are combined with the original training objective as Lfinal = Loriginal + alpha * Laux. The authors evaluate the approach with Wayformer and AutoBots on nuScenes and Argoverse 2, reporting that the auxiliary losses reduce the Offroad metric by about 47% on original scenes and 37% on attacked scenes, improve Direction and Diversity metrics, and broadly maintain minADE/minFDE. The paper also includes qualitative examples, an alpha-sweep study, and a robustness evaluation under the Scene Attack benchmark.

Significance. If the reported results are valid, the proposed losses offer a simple, model-agnostic way to improve road compliance and prediction diversity without architectural changes, which would be practically useful for autonomous driving. The paper provides a public code link and evaluates on two standard datasets with two baselines. However, the significance is substantially tempered by two issues: the evaluation metrics appear to be computed from the same formulas as the training losses, and the main comparison is confounded by an asymmetric fine-tuning protocol. The non-circular evidence, namely that minADE/minFDE are not severely degraded, is credible but reported without error bars or significance testing. Thus the headline safety and robustness claims need additional support before the paper's central contribution can be accepted as established.

major comments (4)
  1. [Section IV-A, Table I, Eqs. (2)-(4)] The evaluation metrics called Offroad, Direction Error, and Diversity in Table I appear to be numerically identical to the training losses defined in Eqs. (2)-(4), yet their definitions are never explicitly given in Section IV-A. If they are the same quantities, then the reported improvements on these metrics are partly by construction: the optimizer is directly minimizing the evaluation yardstick. The manuscript should state the exact metric formulas and, more importantly, provide evidence from independent metrics that are not minimized during training, or explicitly frame the results as an optimization check rather than as independent safety evaluation.
  2. [Section IV-A, training strategy] The training protocol is asymmetric across rows of Table I. The Offroad and Direction losses are trained from scratch, while the Diversity loss is applied by fine-tuning the baseline for 10 epochs; the '+All' model is also fine-tuned. The baseline row receives no equivalent fine-tuning. Consequently, the gains in the '+Diversity' and '+All' rows (e.g., AutoBots nuScenes Diversity from 64 to 93, Wayformer Argoverse2 Direction from 4.24 to 2.54 with '+All') could be due to the additional optimization steps alone rather than to the proposed losses. The causal claim that the proposed losses improve all quality metrics requires a fair comparison with a baseline that is fine-tuned for the same number of epochs under the same schedule.
  3. [Section IV-B, Table I, Fig. 4] All quantitative results are reported as single numbers without error bars, number of seeds, or significance tests. In addition, the alpha hyperparameter is selected from the same trade-off curves shown in Fig. 4 that are used to report the final model, which introduces selection bias. The authors should either fix the alpha selection rule a priori, use a validation split for selection, or report variability across multiple training runs. Without this, it is not possible to assess whether differences such as Wayformer nuScenes minADE from 1.08 to 1.13 or AutoBots Argoverse2 minADE from 0.92 to 0.95 are meaningful.
  4. [Section IV-E and Abstract] The abstract claims an average 37% reduction in offroad errors on attacked scenes, but Table II does not directly reproduce this number for either '+Offroad' or '+All'. For '+All', the attacked-scene reductions range from 15.1% (Wayformer Argoverse2) to 34.1% (AutoBots nuScenes), and for '+Offroad' they range from 15.1% (Wayformer nuScenes) to 59.4% (Wayformer Argoverse2). The manuscript should specify exactly which rows are averaged, how the average is computed, and should report the per-scenario or per-attack-type variability. Similar precision is needed for the 47% original-scene claim.
minor comments (6)
  1. [Section III-A, Eq. (3)] The heading angle gamma_i^t of a predicted trajectory point is used in Eq. (3), but the manuscript never defines how gamma_i^t is computed from the predicted positions. A short formula or reference would remove ambiguity.
  2. [Section III-C.3, Eq. (4)] The double sum in Eq. (4) includes the term i = j, whose pairwise distance is zero but whose indicator functions are still evaluated, and the notation applies 1(j) but not 1(i) in the summand. The formula should be cleaned up, for example by summing over i < j and explicitly applying both feasibility indicators.
  3. [Section IV-A] The text says 'we introduce three novel measures' for evaluation, but no formulas are given for the Offroad, Direction Error, and Diversity metrics. The authors should state explicitly whether these metrics are the same functions as Eqs. (2)-(4) or whether they use different parameters, since this is central to interpreting Table I.
  4. [Fig. 4] The x-axis labels 'Offroad', 'Direction Consistency Error', and 'Diversity' should be clarified as metric names; if they are identical to the loss definitions, the figure should say so, and if not, the metric definitions should be given in the caption.
  5. [Section II, Related Work] The discussion of prior offroad losses in [25]-[29] correctly notes that they rely on rasterized masks, but the claim of novelty should be stated more carefully: the vectorized formulation is a useful engineering contribution, while the concept of an offroad loss is already present in the cited literature.
  6. [Section III-C.1, Eq. (2)] The margin m in Eq. (2) is introduced but its value is never reported in the experiments. Since the margin affects both the loss and the resulting Offroad metric, the chosen values for m, m_d, and m_theta should be listed in the experimental setup.

Circularity Check

1 steps flagged · score 6.0 of 10

Evaluation metrics duplicate the training losses, so the headline Offroad/Direction/Diversity gains are partly by construction.

  1. self definitional [Abstract; Section III-C Eqs. (2)-(4); Section IV-A (evaluation metrics)]
    "These loss functions not only improve model training but can also serve as metrics for evaluating the realism and diversity of trajectory predictions. ... In terms of evaluation metrics, we introduce three novel measures — Offroad, Direction Error, and Diversity — to assess specific aspects of prediction quality."

    Equation (2) defines the Offroad Loss, Eq. (3) defines Direction Consistency, and Eq. (4) defines Mode Diversity. The paper then reports Offroad, Direction Error, and Diversity as the evaluation metrics, using the same signed-distance, centerline-matching, and pairwise-distance formulas. Because the model is trained to minimize these very quantities (via Lfinal = Loriginal + αLaux), the large reductions in Table I's Offroad, Direction, and Diversity columns are the optimizer minimizing the evaluation metric itself. For example, the claim that applying the Offroad loss 'cuts the offroad metric nearly in half' is partly tautological: the model is trained with Eq. (2) and evaluated with the same Offroad formula.

full rationale

The paper's central empirical claim is that three auxiliary losses improve quality metrics. The main circular step is that the three headline quality metrics are identical to the three proposed losses: the Abstract explicitly says the losses 'can also serve as metrics,' and Section IV-A confirms that Offroad, Direction Error, and Diversity are the evaluation measures. Thus Table I's dramatic improvements on those columns are partly by construction. This is not a case of pure self-citation or imported uniqueness; the loss formulas themselves are the content, and the identity is explicit rather than hidden. The less-circular portions are the standard accuracy metrics (minADE, minFDE, MR), which are not optimized by the auxiliary losses and do not collapse, and the attacked-scene robustness results, which evaluate on modified inputs not seen during training. However, even those attacked-scene Offroad numbers use the same Offroad-loss formula, so the safety-improvement claim is only partially independent. The fine-tuning asymmetry between the baseline and '+All' rows is a confound, but it is a fairness issue rather than circularity. Overall, some of the reported improvement reduces by construction to the training objective, but the paper retains independent content, giving a score of 6.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The method is empirical and rests on hyperparameters (margins and loss weights) that are not fully specified, plus annotation assumptions about drivable-area polygons and centerline yaw. No new physical entities are introduced. The most consequential ledger item is the assumption that the evaluation metrics, which mirror the training losses, measure safety; this drives the circularity burden.

free parameters (5)
  • Offroad margin m = not reported
    In Eq 2, m controls the buffer distance from the drivable-area boundary before penalty begins; its value is not stated.
  • Direction distance margin m_d = not reported
    In Eq 3, m_d is the allowable distance threshold before the centerline matching penalty accrues; value not stated.
  • Direction angle margin m_theta = not reported
    In Eq 3, m_theta is the allowable heading difference before penalty accrues; value not stated.
  • Auxiliary loss weight alpha = selected per model/dataset from tuning curves (Fig 4)
    The total loss is L_original + alpha * L_aux; the paper chooses alpha values per configuration and calls the resulting models selected models, which introduces selection risk.
  • Weights for combining three losses in +All = not reported
    The total auxiliary loss is a weighted sum derived from individual alpha values and adjusted for optimal performance; exact weights are not given.
assumptions (5)
  • domain assumption The drivable area can be represented as a union of closed polygons with accurate boundaries.
    Section III-A defines the input drivable area Omega as a union of polygons; if the polygons are noisy or outdated, both loss and metric are distorted.
  • domain assumption Road centerlines with yaw values are available and reliable for all scenes.
    Section III-A includes centerlines c with position and yaw; the Direction loss depends on correct direction annotations.
  • standard math Ray casting and signed distance computation are correctly implemented on GPU.
    Section III-C1 uses the point-in-polygon ray casting algorithm from [32]; correctness is standard but implementation bugs would affect all results.
  • ad hoc to paper Auxiliary losses applied to all modes will not harm the best-mode accuracy.
    The paper's method assumes that supervising all modes with scene-compliance penalties improves safety without degrading minADE; this is an empirical assumption tested but not derived.
  • ad hoc to paper The evaluation metrics are valid proxies for safety and realism.
    The Offroad, Direction, and Diversity metrics used in Table I appear to be the same formulas as the training losses; the paper assumes these quantities capture real-world safety.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Multi-Loss Strategy for Vehicle Trajectory Prediction: Combining Off-Road, Diversity, and Directional Consistency Losses." pith.science (2026). https://pith.science/paper/W73RMALZ

@misc{pith2026241119747,
  author       = {Pith},
  title        = {Pith review of: A Multi-Loss Strategy for Vehicle Trajectory Prediction: Combining Off-Road, Diversity, and Directional Consistency Losses},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/W73RMALZ}},
  note         = {Machine review of arXiv:2411.19747}
}
read the original abstract

Trajectory prediction is essential for the safety and efficiency of planning in autonomous vehicles. However, current models often fail to fully capture complex traffic rules and the complete range of potential vehicle movements. Addressing these limitations, this study introduces three novel loss functions: Offroad Loss, Direction Consistency Error, and Diversity Loss. These functions are designed to keep predicted paths within driving area boundaries, aligned with traffic directions, and cover a wider variety of plausible driving scenarios. As all prediction modes should adhere to road rules and conditions, this work overcomes the shortcomings of traditional "winner takes all" training methods by applying the loss functions to all prediction modes. These loss functions not only improve model training but can also serve as metrics for evaluating the realism and diversity of trajectory predictions. Extensive validation on the nuScenes and Argoverse 2 datasets with leading baseline models demonstrates that our approach not only maintains accuracy but significantly improves safety and robustness, reducing offroad errors on average by 47% on original and by 37% on attacked scenes. This work sets a new benchmark for trajectory prediction in autonomous driving, offering substantial improvements in navigating complex environments. Our code is available at https://github.com/vita-epfl/stay-on-track .

Figures

Figures reproduced from arXiv: 2411.19747 by the authors.

Figure 1
Figure 1. Trajectory predictions by Wayformer [1], a state-of [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. An illustration for our proposed loss functions. The colors in (a) show the Offroad Loss values for areas around the [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Comparative visualization of model predictions on the Argoverse 2 dataset using Wayformer as the baseline. Each [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Performance impact of integrating our loss functions on Wayformer’s minADE across the Argoverse 2 dataset. The blue [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 30 canonical work pages

  1. [1]

    Wayformer: Motion forecasting via simple & efficient attention networks,

    N. Nayakanti, R. Al-Rfou, A. Zhou, K. Goel, K. S. Refaat, and B. Sapp, “Wayformer: Motion forecasting via simple & efficient attention networks,” 2023 IEEE International Conference on Robotics and Automation (ICRA) , pp. 2980–2987, 2022. [Online]. Available: https://api.semanticscholar.org/CorpusID:250493056

    work page 2023

  2. [2]

    Safety-aware motion prediction with unseen vehicles for autonomous driving,

    X. Ren, T. Yang, L. E. Li, A. Alahi, and Q. Chen, “Safety-aware motion prediction with unseen vehicles for autonomous driving,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2021

    work page 2021

  3. [3]

    Prank: motion prediction based on ranking,

    Y . Biktairov, M. Stebelev, I. Rudenko, O. Shliazhko, and B. Yangel, “Prank: motion prediction based on ranking,” Advances in Neural Information Processing Systems (NeurIPS) , 2020

    work page 2020

  4. [4]

    Multipath: Multiple probabilistic anchor trajectory hypotheses for behavior prediction,

    Y . Chai, B. Sapp, M. Bansal, and D. Anguelov, “Multipath: Multiple probabilistic anchor trajectory hypotheses for behavior prediction,” Conference on Robot Learning , 2019

    work page 2019

  5. [5]

    Rules of the road: Predicting driving behavior with a convolutional model of semantic interactions,

    J. Hong, B. Sapp, and J. Philbin, “Rules of the road: Predicting driving behavior with a convolutional model of semantic interactions,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2019

    work page 2019

  6. [6]

    Intentnet: Learning to predict intention from raw sensor data,

    S. Casas, W. Luo, and R. Urtasun, “Intentnet: Learning to predict intention from raw sensor data,” in Conference on Robot Learning . PMLR, 2018

    work page 2018

  7. [7]

    Learning lane graph representations for motion forecasting,

    M. Liang, B. Yang, R. Hu, Y . Chen, R. Liao, S. Feng, and R. Urta- sun, “Learning lane graph representations for motion forecasting,” in European Conference on Computer Vision (ECCV) . Springer, 2020

    work page 2020

  8. [8]

    Vectornet: Encoding hd maps and agent dynamics from vectorized representation,

    J. Gao, C. Sun, H. Zhao, Y . Shen, D. Anguelov, C. Li, and C. Schmid, “Vectornet: Encoding hd maps and agent dynamics from vectorized representation,” in Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition (CVPR) , 2020

    work page 2020

Show all 36 references
  1. [9]

    Learning heterogeneous interaction strengths by trajectory prediction with graph neural network,

    S. Ha and H. Jeong, “Learning heterogeneous interaction strengths by trajectory prediction with graph neural network,” in The Eleventh International Conference on Learning Representations , 2023. [Online]. Available: https://openreview.net/forum?id=qU6NIcpaSi-

    2023

  2. [10]

    Graph-based spatial transformer with memory replay for multi-future pedestrian trajectory prediction,

    L. Li, M. Pagnucco, and Y . Song, “Graph-based spatial transformer with memory replay for multi-future pedestrian trajectory prediction,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 2231–2241

    2022

  3. [11]

    Diverse and admissible trajectory forecasting through multimodal context understanding,

    S. H. Park, G. Lee, J. Seo, M. Bhat, M. Kang, J. Francis, A. Jadhav, P. P. Liang, and L.-P. Morency, “Diverse and admissible trajectory forecasting through multimodal context understanding,” in European Conference on Computer Vision (ECCV) . Springer, 2020

    2020

  4. [12]

    Goal-driven self-attentive recurrent networks for trajectory prediction,

    L. Chiara, P. Coscia, S. Das, S. Calderara, R. Cucchiara, and L. Ballan, “Goal-driven self-attentive recurrent networks for trajectory prediction,” in 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) . Los Alamitos, CA, USA: IEEE Computer S...

    2022

  5. [13]

    Vehicle trajectory prediction using LSTMs with spatial-temporal attention mechanisms,

    L. Lin, W. Li, H. Bi, and L. Qin, “Vehicle trajectory prediction using LSTMs with spatial-temporal attention mechanisms,” IEEE Intelligent Transportation Systems Magazine, vol. 14, no. 2, pp. 197—-208, 2022

    2022

  6. [14]

    Vehicle trajectory prediction based on lstm recurrent neural networks,

    A. Ip, L. Irio, and R. Oliveira, “Vehicle trajectory prediction based on lstm recurrent neural networks,” in 2021 IEEE 93rd V ehicular Technology Conference (VTC2021-Spring) , 2021, pp. 1–5

    2021

  7. [15]

    Latent variable sequential set transformers for joint multi-agent motion prediction,

    R. Girgis, F. Golemo, F. Codevilla, M. Weiss, J. A. D’Souza, S. E. Kahou, F. Heide, and C. Pal, “Latent variable sequential set transformers for joint multi-agent motion prediction,” in International Conference on Learning Representations , 2022. [Online]. Available: https://o...

    2022

  8. [16]

    Laformer: Trajectory prediction for autonomous driving with lane-aware scene constraints,

    M. Liu, H. Cheng, L. Chen, H. Broszio, J. Li, R. Zhao, M. Sester, and 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

    2024

  9. [17]

    Motion transformer with global intention localization and local movement refinement,

    S. Shi, L. Jiang, D. Dai, and B. Schiele, “Motion transformer with global intention localization and local movement refinement,” Ad- vances in Neural Information Processing Systems , 2022

    2022

  10. [18]

    Gameformer: Game-theoretic modeling and learning of transformer-based interactive prediction and planning for autonomous driving,

    Z. Huang, H. Liu, and C. Lv, “Gameformer: Game-theoretic modeling and learning of transformer-based interactive prediction and planning for autonomous driving,” in Proceedings of the IEEE/CVF Interna- tional Conference on Computer Vision (ICCV) , October 2023, pp. 3903–3913

    2023

  11. [19]

    Multimodal Trajectory Prediction Conditioned on Lane-Graph Traversals,

    N. Deo, E. Wolff, and O. Beijbom, “Multimodal Trajectory Prediction Conditioned on Lane-Graph Traversals,” in Proceedings of the 5th Conference on Robot Learning . PMLR, Jan. 2022, pp. 203–212, iSSN: 2640-3498. [Online]. Available: https://proceedings.mlr.press/ v164/deo22a.html

    2022

  12. [20]

    Gohome: Graph-oriented heatmap output for future motion estimation,

    T. Gilles, S. Sabatini, D. V . Tsishkou, B. Stanciulescu, and F. Moutarde, “Gohome: Graph-oriented heatmap output for future motion estimation,” 2022 International Conference on Robotics and Automation (ICRA) , pp. 9107–9114, 2021. [Online]. Available: https://api.semanticscho...

    2022

  13. [21]

    Densetnt: End-to-end trajectory pre- diction from dense goal sets,

    J. Gu, C. Sun, and H. Zhao, “Densetnt: End-to-end trajectory pre- diction from dense goal sets,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2021, pp. 15 303– 15 312

    2021

  14. [22]

    Ltp: Lane-based trajectory prediction for autonomous driving,

    J. Wang, T. Ye, Z. Gu, and J. Chen, “Ltp: Lane-based trajectory prediction for autonomous driving,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2022, pp. 17 134–17 142

    2022

  15. [23]

    Diverse multiple trajectory prediction using a two-stage prediction network trained with lane loss,

    S. Kim, H. Jeon, J. W. Choi, and D. Kum, “Diverse multiple trajectory prediction using a two-stage prediction network trained with lane loss,” IEEE Robotics and Automation Letters , vol. 8, no. 4, pp. 2038–2045, 2023

    2023

  16. [24]

    Scene compliant trajectory forecast with agent-centric spatio-temporal grids,

    D. Ridel, N. Deo, D. Wolf, and M. Trivedi, “Scene compliant trajectory forecast with agent-centric spatio-temporal grids,” IEEE Robotics and Automation Letters, vol. 5, no. 2, pp. 2816–2823, 2020

    2020

  17. [25]

    Improving movement prediction of traffic actors using off-road loss and bias mitigation,

    M. Niedoba, H. Cui, K. Luo, D. Hegde, F.-C. Chou, and N. Djuric, “Improving movement prediction of traffic actors using off-road loss and bias mitigation,” in Workshop on’Machine Learning for Autonomous Driving’at Conference on Neural Information Processing Systems, 2019

    2019

  18. [26]

    Motion prediction us- ing trajectory sets and self-driving domain knowledge,

    F. A. Boulton, E. C. Grigore, and E. M. Wolff, “Motion prediction us- ing trajectory sets and self-driving domain knowledge,” arXiv preprint arXiv:2006.04767, 2020

    2006 arXiv

  19. [27]

    Multi-head attention with joint agent-map representation for trajectory prediction in autonomous driving,

    K. Messaoud, N. Deo, M. M. Trivedi, and F. Nashashibi, “Multi-head attention with joint agent-map representation for trajectory prediction in autonomous driving,” arXiv preprint arXiv:2005.02545 , 2020

    2005 arXiv

  20. [28]

    Argoverse: 3d tracking and forecasting with rich maps,

    M.-F. Chang, J. W. Lambert, P. Sangkloy, J. Singh, S. Bak, A. Hartnett, D. Wang, P. Carr, S. Lucey, D. Ramanan, and J. Hays, “Argoverse: 3d tracking and forecasting with rich maps,” in Conference on Computer Vision and Pattern Recognition (CVPR) , 2019

    2019

  21. [29]

    Ellipse loss for scene-compliant motion prediction,

    H. Cui, H. Shajari, S. Yalamanchi, and N. Djuric, “Ellipse loss for scene-compliant motion prediction,” in 2021 IEEE International Conference on Robotics and Automation (ICRA) , 2021, pp. 8558– 8564

    2021

  22. [30]

    Trajectory prediction in au- tonomous driving with a lane heading auxiliary loss,

    R. Greer, N. Deo, and M. Trivedi, “Trajectory prediction in au- tonomous driving with a lane heading auxiliary loss,” IEEE Robotics and Automation Letters , vol. PP, pp. 1–1, 03 2021

    2021

  23. [31]

    The importance of prior knowledge in precise multimodal prediction,

    S. Casas, C. Gulino, S. Suo, and R. Urtasun, “The importance of prior knowledge in precise multimodal prediction,” 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pp. 2295–2302, 2020. [Online]. Available: https: //api.semanticscholar.org/Corpu...

    2020

  24. [32]

    O’Rourke, Computational Geometry in C , 2nd ed

    J. O’Rourke, Computational Geometry in C , 2nd ed. Cambridge, UK: Cambridge University Press, 1998

    1998

  25. [33]

    Vehicle trajectory prediction works, but not everywhere,

    M. Bahari, S. Saadatnejad, A. Rahimi, M. Shaverdikondori, A.- H. Shahidzadeh, S.-M. Moosavi-Dezfooli, and A. Alahi, “Vehicle trajectory prediction works, but not everywhere,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022

    2022

  26. [34]

    UniTraj: A Unified Framework for Scalable Vehicle Trajectory Prediction,

    L. Feng, M. Bahari, K. M. B. Amor, E. Zablocki, M. Cord, and A. Alahi, “UniTraj: A Unified Framework for Scalable Vehicle Trajectory Prediction,” Aug. 2024, arXiv:2403.15098 [cs]. [Online]. Available: http://arxiv.org/abs/2403.15098

    2024 arXiv

  27. [35]

    nuscenes: A multimodal dataset for autonomous driving,

    H. Caesar, V . Bankiti, A. H. Lang, S. V ora, V . E. Liong, Q. Xu, A. Krishnan, Y . Pan, G. Baldan, and O. Beijbom, “nuscenes: A multimodal dataset for autonomous driving,” 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 11 618–11 628, 2019. [On...

    2020

  28. [36]

    Argoverse 2: Next generation datasets for self-driving perception and forecasting,

    B. Wilson, W. Qi, T. Agarwal, J. Lambert, J. Singh, S. Khandelwal, B. Pan, R. Kumar, A. Hartnett, J. K. Pontes, D. Ramanan, P. Carr, and J. Hays, “Argoverse 2: Next generation datasets for self-driving perception and forecasting,” in Proceedings of the Neural Information Proce...

    2021

Pith tools

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