Pith. sign in

REVIEW 4 major objections 7 minor 37 references

LHPF: Look back the History and Plan for the Future in Autonomous Driving

T0 review · 4 major / 7 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 purely learning-based driving planner reports its first win over the expert baseline by remembering its own past plans.

desk verdict A credible temporal-ensembling extension to imitation planners with a solid gain over its backbone, but the 'surpasses the expert' claim rides on a 0.01-point margin with no error bars. read the letter →

arxiv 2411.17253 v1 pith:BDDVPG3O submitted 2024-11-26 cs.RO cs.CV

Sheng Wang , Yao Tian , Xiaodong Mei , Ge Sun , Jie Cheng , Fulong Ma , Pedro V. Sander , Junwei Liang This is my paper · ORCID
classification cs.ROcs.CV
keywords imitationlearningautonomousdrivingplanningclosed-loophistoricalintentionaggregationNuPlanbenchmarkspatio-temporalquerycomfortlosstrajectory
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

LHPF is an imitation-learning planner that tries to fix a specific failure: conventional planners treat each frame's plan in isolation, so once small errors accumulate the vehicle forgets its own earlier intention (for example, an overtake) and drifts off course. The paper's central claim is that feeding the planner its own historical planning embeddings, pooled over past frames and fused with current spatial queries, restores temporal continuity and, together with a comfort auxiliary loss, improves closed-loop performance. On the NuPlan Val14 reactive benchmark the method scores 81.25, above the expert log-replay baseline of 81.24, which the authors report as the first time a purely learning-based planner has surpassed the expert. If this holds, imitation planners can use their own past decisions as a memory mechanism rather than re-planning from scratch at every step, and the module can be fine-tuned onto an existing backbone in about five epochs.

What carries the argument

The load-bearing component is the historical intention aggregation module. At each timestep the query-based decoder's latent output is stored as a planning embedding; over the past observation window these embeddings are pooled and fused with the current spatial queries (lateral queries from the reference-line polyline encoder and longitudinal queries from learnable embeddings) to form spatio-temporal queries. A spatio-temporal decoder with lateral self-attention, longitudinal self-attention, and query-to-scene cross-attention then produces the final trajectory and score. The comfort auxiliary loss, computed from max-exceedance penalties over longitudinal and lateral acceleration, yaw rate, yaw acceleration, and jerk, is what lets the method recover comfort lost by the more aggressive driving style that historical intentions induce.

What would settle it

Re-run the NuPlan Val14 reactive closed-loop evaluation for LHPF multiple times with different random seeds and/or fine-tuning checkpoints using the official simulator and scoring code. If the distribution of LHPF scores overlaps or falls below the expert log-replay score of 81.24, in particular if 81.25 is not reproducible, the central claim of surpassing the expert fails. A cheaper check is to report per-scenario score variance across the 1,090 scenarios and see whether a 0.01 gap is within the standard error of the mean.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that a planner's previously generated plans are a usable and useful signal for its current decision. LHPF collects the latent planning embeddings from earlier timesteps into a historical intention pool, combines them with the current spatial reference-line queries to form spatio-temporal queries, and decodes the final trajectory from those queries with a query-to-scene transformer decoder. A comfort loss that penalizes trajectories exceeding human-like acceleration, jerk, and yaw limits is added during fine-tuning. The result on the NuPlan Val14 reactive closed-loop benchmark is a total score of 81.25, the highest among purely learning-based planners and, against the expert log-replay baseline of 81.24, a reported first for a purely learning-based method beating the expert. The same historical-intention module, attached to a different backbone and evaluated on CommonRoad, also improves completion distance and reduces side collisions.

Load-bearing premise

The paper's headline result rests on a single evaluation in which LHPF's reactive score of 81.25 exceeds the expert log-replay score of 81.24 by 0.01 points; the paper gives no error bars, multiple seeds, or significance test, so the claim that a purely learning-based planner has beaten the expert assumes that 0.01-point gap is not evaluation noise.

Editorial extensions

If this is right

  • If correct, imitation-learning planners no longer need to re-derive intent from scratch every frame; keeping a memory of one's own plan is a cheap way to resist covariate shift and error accumulation.
  • The module is a plug-in: starting from a pretrained PLUTO checkpoint, five epochs of fine-tuning on the spatio-temporal decoder yields the reported gain of roughly 2.5 reactive-score points over PLUTO without post-processing.
  • On CommonRoad with a PGP backbone, adding the historical-intention module improves average completion distance from 25.18 to 26.46 and cuts side collisions from 16 to 9, suggesting the mechanism transfers across planners and environments.
  • Because the improvement comes mainly through progress and assertiveness rather than safety metrics, the method changes the trade-off frontier: more intent-consistent driving at the cost of slightly more collisions (95.82 versus PLUTO's 97.43).
  • The best fusion is summation rather than attention, and a 10-frame historical interval beats both the 20-frame setting (no history) and the 1-frame setting (dense, noisy history).

Reading between the lines

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

  • Beyond the paper: the 0.01-point margin over expert log-replay is within what a single run of a stochastic planner could produce by chance; without repeated seeds or error bars, the 'first to beat the expert' headline should be read as provisional until replicated.
  • Beyond the paper: historical planning embeddings may act as a temporal regularizer or low-pass filter on intentions; if so, similar gains could come from simpler mechanisms such as exponential moving averages over plan embeddings, which is a direct testable extension.
  • Beyond the paper: the fact that denser history (interval 1) hurts suggests the module is not simply 'more memory is better'; finding the right temporal resolution may matter more than the pooling operator, an effect the paper reports but does not explain.
  • Beyond the paper: applying the same historical-intention module to end-to-end planners that consume raw sensor data, where distribution shift is larger, would test whether the memory mechanism's benefit grows with the degree of covariate shift.
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

4 major / 7 minor

Summary. The paper proposes LHPF, an imitation-learning planner built on the PLUTO backbone. It augments a query-based decoder with a historical intention pool: planning embeddings from past frames are stored and fused with current spatial queries, and a comfort auxiliary loss is added during a short fine-tuning phase. The authors report a reactive closed-loop R-Score of 81.25 on the NuPlan Val14 benchmark, which they claim is the first purely learning-based planner to surpass the expert log-replay baseline of 81.24. They also report experiments on CommonRoad with a PGP backbone, and ablations on historical interval, fusion method, and training epochs.

Significance. If the headline result is reproducible, the paper makes a useful empirical contribution: a lightweight plug-in module that improves an existing strong planner without modifying its backbone, requires only a few fine-tuning epochs, and improves closed-loop progress. The cross-backbone experiment (PGP on CommonRoad) strengthens the claim that the historical-intention mechanism generalizes. However, the central milestone claim—that LHPF is the first learning-based planner to exceed expert performance—rests on a 0.01-point gap from a single evaluation, with no variance, multiple seeds, or significance testing. The contribution is therefore plausible but not yet established at the level the abstract and introduction assert. The paper also contains an internal numerical inconsistency in Section 5.1 that directly affects confidence in the reported scores. I would support publication after the authors address the statistical support for the headline claim and correct the reporting inconsistencies.

major comments (4)
  1. [Section 5.1, Table 1, Abstract] The headline claim that LHPF is the first purely learning-based planner to surpass the expert rests on a 0.01-point margin in R-Score (81.25 vs. 81.24). The expert baseline is a fixed log-replay score, but LHPF is a trained neural planner evaluated once in a stochastic simulator; no seeds, repeated runs, error bars, or significance tests are reported. A single training run or evaluation can plausibly shift the score by more than 0.01, which would erase the claimed milestone. Please report the variance over multiple training seeds and evaluation runs, or explicitly soften the claim to 'matched or marginally exceeded in this evaluation' until statistical support is available.
  2. [Section 5.1, paragraph after Table 1] The text states that 'Compared to the baseline method, Pluto, the integration of historical planning information results in an improvement of the Progress metric from 78.59 to 79.21,' but Table 1 reports PLUTO Progress as 75.08 and LHPF Progress as 78.59. Table 2 also lists the scratch baseline Progress as 75.08. The reported 78.59 value appears to be LHPF's own Progress, not PLUTO's, and the 79.21 value appears nowhere in the tables. This internal inconsistency in the exact numbers supporting the narrative must be corrected, and the corrected values should be checked against the evaluation logs.
  3. [Section 5.4, Table 2] The ablation study states that 'we use 10 scenarios per type from the nuPlan val14 dataset' for the ablation, which is roughly 140 of the 1090 scenarios in Val14. Yet Table 2 lists the same R-Score values (80.53, 81.25, etc.) that appear to be full-benchmark scores, and the baseline row is 78.76, identical to the full-benchmark PLUTO score in Table 1. Please clarify whether Table 2 reports full Val14 results or subset results. If it is a subset, the choice of the best configuration (summation fusion plus comfort loss) should be validated on the full benchmark, and the subset results need error bars because random subset selection can change scores by more than the 0.01-point margin that is central to the paper.
  4. [Table 3, CommonRoad results] The text says L-PGP 'substantially improves collision metrics' relative to PGP, but the numbers show FCT at 26 vs. 27 and SCT at 9 vs. 16, with JERK increasing from 41.89 to 45.67. The only large improvement is in SCT, and it is measured as a count over 54 scenarios with no measure of variability. Please report per-scenario statistics or another suitable error assessment, and phrase the claim to match the mixed result: improved side collisions at the cost of higher jerk and essentially unchanged front collisions.
minor comments (7)
  1. [Section 2, Related Work] The sentence 'Despite these advances, many methods rely on still suffer from the sim-to-real problem. specifically' is garbled and should be rewritten.
  2. [Section 3.5, Eq. for Qst] The summation notation '0X i=-TH' is unusual and the bounds are ambiguous; please rewrite with standard summation limits, e.g., sum over historical timesteps i = -TH, ..., 0.
  3. [Figure 5 and Section 5.4] The discussion of historical planning intervals and training epochs refers to performance peaks and drops, but Figure 5 appears to lack axis labels and error bars, and no numerical values are given. Adding a table of the plotted values would make the ablation reproducible.
  4. [Table 2] The row label 'Scratch' is confusing because the text identifies the scratch model as PLUTO; renaming the row to 'PLUTO (backbone)' would align with Table 1.
  5. [Table 6] Table 6 lists both 'PLUTO [9] 92.06' and 'PLUTO (w/o post.) [9] 78.76' without explaining that the difference is post-processing; a footnote or column header would prevent reader confusion.
  6. [References] References [26] and [27] both cite UrbanDriver and appear to be duplicate entries; please merge them and update the citation numbers.
  7. [Algorithm 1] There are typos in the supplementary material, e.g., 'aslo' in the inference description; a careful proofread is needed.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation-to-input reduction; minor coauthored citations are used as external published baselines and do not force the headline result.

full rationale

LHPF makes empirical benchmark claims rather than deriving a target quantity from its own definition. The central result (R-Score 81.25 on NuPlan Val14) is measured in a closed-loop simulator against external baselines including Expert Log-Replay. The PLUTO backbone [9] is coauthored by present authors, but it is a published method with an official checkpoint and is used as the starting point and baseline; the comparison is external and falsifiable, so this self-citation is not load-bearing. The historical-intention pool feeds the planner's own previous embeddings into the current query, which is a recurrent design choice, not a logical reduction of the output to the input. The comfort-loss thresholds and historical interval are hand-set or tuned on a subset of the same benchmark; that is a benchmark-overfitting and statistical-robustness concern, especially given the 0.01-point margin over the expert from a single run with no error bars, but it is not circularity by construction. The Section 5.1 Progress discrepancy (78.59 to 79.21 vs. Table 1 values 75.08 and 78.59) is a numerical inconsistency, not a circular step. No equation is defined in terms of the quantity it is supposed to predict, and no fitted parameter is renamed as a prediction. Score 2 reflects only minor, non-load-bearing self-citations to coauthored prior work.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The central claim is empirical and rests on benchmark assumptions and hand-set hyperparameters, not on a mathematical derivation. No new physical entities are introduced.

free parameters (4)
  • Historical planning interval = 10 frames
    Selected via ablation on a Val14 subset (Figure 5); an interval of 1 degrades R-Score below baseline, so the chosen value is load-bearing for the reported result.
  • Fine-tuning epochs = 5
    Selected from Figure 5 as the peak of fine-tuning performance; the reported result depends on stopping at this epoch.
  • Comfort loss dynamic thresholds = See Table 5, e.g., max longitudinal acceleration -4.05 to 2.40
    Hand-set constants in Lcomfort that trade off comfort against progress; no source or fitting procedure is given for these values.
  • Loss weights = 1.0 for all terms
    Chosen without tuning; a reasonable default, but still a hand choice that affects the balance between imitation and comfort losses.
assumptions (3)
  • domain assumption NuPlan Val14 reactive closed-loop score is a valid proxy for planning quality and safety.
    The entire comparison and the 'surpass expert' claim are made on this benchmark; the paper states the reactive score is the primary metric in Section 4.1.
  • domain assumption Historical planning embeddings generated by a frozen PLUTO encoder remain informative when the planner is deployed closed-loop and conditions on its own outputs.
    The mechanism assumes self-conditioning on prior plans improves intention consistency rather than compounding errors; training uses logged frames while closed-loop evaluation uses rollout plans, per Sections 3.4, 3.5, and 4.
  • domain assumption The one-hot projected ground-truth trajectory, denoted tau-hat, is a valid single supervision target.
    Lreg and Lcls are computed against tau-hat projected onto the reference line in Section 3.6; if this projection is ambiguous in intersections or lane changes, the supervision signal changes.

how reviews work

0 comments
Cite this review

Pith. "Pith review of LHPF: Look back the History and Plan for the Future in Autonomous Driving." pith.science (2026). https://pith.science/paper/BDDVPG3O

@misc{pith2026241117253,
  author       = {Pith},
  title        = {Pith review of: LHPF: Look back the History and Plan for the Future in Autonomous Driving},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BDDVPG3O}},
  note         = {Machine review of arXiv:2411.17253}
}
read the original abstract

Decision-making and planning in autonomous driving critically reflect the safety of the system, making effective planning imperative. Current imitation learning-based planning algorithms often merge historical trajectories with present observations to predict future candidate paths. However, these algorithms typically assess the current and historical plans independently, leading to discontinuities in driving intentions and an accumulation of errors with each step in a discontinuous plan. To tackle this challenge, this paper introduces LHPF, an imitation learning planner that integrates historical planning information. Our approach employs a historical intention aggregation module that pools historical planning intentions, which are then combined with a spatial query vector to decode the final planning trajectory. Furthermore, we incorporate a comfort auxiliary task to enhance the human-like quality of the driving behavior. Extensive experiments using both real-world and synthetic data demonstrate that LHPF not only surpasses existing advanced learning-based planners in planning performance but also marks the first instance of a purely learning-based planner outperforming the expert. Additionally, the application of the historical intention aggregation module across various backbones highlights the considerable potential of the proposed method. The code will be made publicly available.

Figures

Figures reproduced from arXiv: 2411.17253 by the authors.

Figure 1
Figure 1. Comparison of imitation learning planners with and [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Planning with Historical Intentions. A stack of planners generates historical planning embeddings at each time step, which are stored in a historical intention pool and combined with reference line queries. Spatio-temporal queries are then aggregated and processed using self and cross attention with scene context as keys and values. Finally, the current planning embedding is passed through a multi-layer perceptron t… view at source ↗
Figure 3
Figure 3. Planning Stream Demonstration. The diagram illus￾trates planning across consecutive frames, with dashed lines high￾lighting significant overlap that reveals the continuity and evolu￾tion of driving behaviors. Q ′ i−1 = SelfAttn(Qi−1, dim = 0), Qˆ i−1 = SelfAttn(Q ′ i−1 , dim = 1), Qi = CrossAttn(Qˆ i−1, Eenc, Eenc). (1) Here, SelfAttn(X, dim = i) denotes self-attention applied across the i-th dimension of X, and Cro… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Demonstration of Reactive Closed-Loop Simulation in Five Representative Scenarios. The rectangles represent the vehicles, with the orange rectangle denoting the autonomous ego vehicle. Additional legends are provided in (a). overtake but quickly abandons this maneuver …
Figure 5
Figure 5. Figure 5: Planning performance comparison with different histor [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: More visualization comparisons of Reactive Closed-Loop Simulation on Nuplan V14 validation set. The rectangles represent the vehicles, with the orange rectangle denoting the autonomous ego vehicle [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]
Figure 7
Figure 7. Figure 7: More visualization comparisons of Reactive Closed-Loop Simulation on Nuplan V14 validation set. The rectangles represent the vehicles, with the orange rectangle denoting the autonomous ego vehicle [PITH_FULL_IMAGE:figures/full_fig_p015_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

37 extracted references · 30 canonical work pages

  1. [1]

    Commonroad: Composable benchmarks for motion plan- ning on roads

    Matthias Althoff, Markus Koschi, and Stefanie Manzinger. Commonroad: Composable benchmarks for motion plan- ning on roads. In 2017 IEEE Intelligent Vehicles Symposium (IV), pages 719–726, 2017. 7, 13

    work page 2017

  2. [2]

    Chauf- feurnet: Learning to drive by imitating the best and synthe- sizing the worst, 2018

    Mayank Bansal, Alex Krizhevsky, and Abhijit Ogale. Chauf- feurnet: Learning to drive by imitating the best and synthe- sizing the worst, 2018. 2

    work page 2018

  3. [3]

    nuplan: A closed-loop ml-based plan- ning benchmark for autonomous vehicles

    Holger Caesar, Juraj Kabzan, Kok Seang Tan, Whye Kit Fong, Eric Wolff, Alex Lang, Luke Fletcher, Oscar Beijbom, and Sammy Omari. nuplan: A closed-loop ml-based plan- ning benchmark for autonomous vehicles. arXiv preprint arXiv:2106.11810, 2021. 1, 5, 6, 12

    arXiv 2021

  4. [4]

    Learning from all vehi- cles

    Dian Chen and Philipp Krähenbühl. Learning from all vehi- cles. In Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition , pages 17222–17231,

    work page

  5. [5]

    Learning by cheating

    Dian Chen, Brady Zhou, Vladlen Koltun, and Philipp Krähenbühl. Learning by cheating. In Conference on Robot Learning, pages 66–75. PMLR, 2020. 1, 2

    work page 2020

  6. [6]

    Ir-stp: Enhancing autonomous driving with interaction reasoning in spatio- temporal planning

    Yingbing Chen, Jie Cheng, Lu Gan, Sheng Wang, Hongji Liu, Xiaodong Mei, and Ming Liu. Ir-stp: Enhancing autonomous driving with interaction reasoning in spatio- temporal planning. IEEE Transactions on Intelligent Trans- portation Systems, pages 1–13, 2024. 7, 8, 13

    work page 2024

  7. [7]

    Mpnp: Multi-policy neural planner for urban driving

    Jie Cheng, Ren Xin, Sheng Wang, and Ming Liu. Mpnp: Multi-policy neural planner for urban driving. In 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pages 10549–10554. IEEE, 2022. 1, 2

    work page 2022

  8. [8]

    Rethinking imitation-based planner for autonomous driving, 2023

    Jie Cheng, Yingbing Chen, Xiaodong Mei, Bowen Yang, Bo Li, and Ming Liu. Rethinking imitation-based planner for autonomous driving, 2023. 1, 2, 3, 6, 12

    work page 2023

Show all 37 references
  1. [9]

    Pluto: Push- ing the limit of imitation learning-based planning for au- tonomous driving

    Jie Cheng, Yingbing Chen, and Qifeng Chen. Pluto: Push- ing the limit of imitation learning-based planning for au- tonomous driving. arXiv preprint arXiv:2404.14327, 2024. 2, 5, 6, 12

    2024 arXiv

  2. [10]

    Neat: Neural attention fields for end-to-end autonomous driving

    Kashyap Chitta, Aditya Prakash, and Andreas Geiger. Neat: Neural attention fields for end-to-end autonomous driving. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 15793–15803, 2021. 2

    2021

  3. [11]

    Transfuser: Imitation with transformer-based sensor fusion for autonomous driv- ing

    Kashyap Chitta, Aditya Prakash, Bernhard Jaeger, Zehao Yu, Katrin Renz, and Andreas Geiger. Transfuser: Imitation with transformer-based sensor fusion for autonomous driv- ing. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(11):12878–12895, 2022. 1, 2

    2022

  4. [12]

    Exploring the limitations of behavior cloning for autonomous driving

    Felipe Codevilla, Eder Santana, Antonio M López, and Adrien Gaidon. Exploring the limitations of behavior cloning for autonomous driving. In Proceedings of the IEEE/CVF international conference on computer vision , pages 9329–9338, 2019. 1, 2

    2019

  5. [13]

    Multimodal trajectory predictions for autonomous driving using deep convolutional networks

    Henggang Cui, Vladan Radosavljevic, Fang-Chieh Chou, Tsung-Han Lin, Thi Nguyen, Tzu-Kuo Huang, Jeff Schnei- der, and Nemanja Djuric. Multimodal trajectory predictions for autonomous driving using deep convolutional networks. In 2019 international conference on robotics and aut...

    2019

  6. [14]

    Parting with misconceptions about learning- based vehicle motion planning

    Daniel Dauner, Marcel Hallgarten, Andreas Geiger, and Kashyap Chitta. Parting with misconceptions about learning- based vehicle motion planning. In Conference on Robot Learning, pages 1268–1281. PMLR, 2023. 1, 2, 5, 6, 12

    2023

  7. [15]

    Multimodal trajectory prediction conditioned on lane-graph traversals

    Nachiket Deo, Eric Wolff, and Oscar Beijbom. Multimodal trajectory prediction conditioned on lane-graph traversals. In 5th Annual Conference on Robot Learning, 2021. 7, 8, 13

    2021

  8. [16]

    Densetnt: End-to-end trajectory prediction from dense goal sets

    Junru Gu, Chen Sun, and Hang Zhao. Densetnt: End-to-end trajectory prediction from dense goal sets. In Proceedings of the IEEE/CVF International Conference on Computer Vi- sion, pages 15303–15312, 2021. 2

    2021

  9. [17]

    From prediction to planning with goal conditioned lane graph traversals

    Marcel Hallgarten, Martin Stoll, and Andreas Zell. From prediction to planning with goal conditioned lane graph traversals. In 2023 IEEE 26th International Conference on Intelligent Transportation Systems (ITSC) , pages 951–958. IEEE, 2023. 1, 6, 12

    2023

  10. [18]

    Multi-modal motion prediction using temporal ensembling with learning-based aggregation, 2024

    Kai-Yin Hong, Chieh-Chih Wang, and Wen-Chieh Lin. Multi-modal motion prediction using temporal ensembling with learning-based aggregation, 2024. 2

    2024

  11. [19]

    St-p3: End-to-end vision-based au- tonomous driving via spatial-temporal feature learning

    Shengchao Hu, Li Chen, Penghao Wu, Hongyang Li, Junchi Yan, and Dacheng Tao. St-p3: End-to-end vision-based au- tonomous driving via spatial-temporal feature learning. In European Conference on Computer Vision, pages 533–549. Springer, 2022. 2

    2022

  12. [20]

    Planning-oriented autonomous driving

    Yihan Hu, Jiazhi Yang, Li Chen, Keyu Li, Chonghao Sima, Xizhou Zhu, Siqi Chai, Senyao Du, Tianwei Lin, Wenhai Wang, et al. Planning-oriented autonomous driving. In Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 17853–17862, 2023. 1, 2

    2023

  13. [21]

    Multi-modal mo- tion prediction with transformer-based neural network for autonomous driving

    Zhiyu Huang, Xiaoyu Mo, and Chen Lv. Multi-modal mo- tion prediction with transformer-based neural network for autonomous driving. In 2022 International Conference on Robotics and Automation (ICRA) , pages 2605–2611. IEEE,

    2022

  14. [22]

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

    Zhiyu Huang, Haochen Liu, and Chen Lv. Gameformer: Game-theoretic modeling and learning of transformer-based interactive prediction and planning for autonomous driving. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 3903–3913, 2023. 1, 2

    2023

  15. [23]

    Dtpp: Differentiable joint con- ditional prediction and cost evaluation for tree policy plan- ning in autonomous driving, 2024

    Zhiyu Huang, Peter Karkus, Boris Ivanovic, Yuxiao Chen, Marco Pavone, and Chen Lv. Dtpp: Differentiable joint con- ditional prediction and cost evaluation for tree policy plan- ning in autonomous driving, 2024. 2

    2024

  16. [24]

    Vad: Vectorized scene representa- tion for efficient autonomous driving

    Bo Jiang, Shaoyu Chen, Qing Xu, Bencheng Liao, Jiajie Chen, Helong Zhou, Qian Zhang, Wenyu Liu, Chang Huang, and Xinggang Wang. Vad: Vectorized scene representa- tion for efficient autonomous driving. In Proceedings of the IEEE/CVF International Conference on Computer Vision ,...

    2023

  17. [25]

    Lapred: Lane-aware prediction of multi- modal future trajectories of dynamic agents

    ByeoungDo Kim, Seong Hyeon Park, Seokhwan Lee, Elbek Khoshimjonov, Dongsuk Kum, Junsoo Kim, Jeong Soo Kim, and Jun Won Choi. Lapred: Lane-aware prediction of multi- modal future trajectories of dynamic agents. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pa...

    2021

  18. [27]

    Urban driver: Learning to drive from real-world demonstrations using policy gradients

    Oliver Scheel, Luca Bergamini, Maciej Wolczyk, Bła ˙zej Osi´nski, and Peter Ondruska. Urban driver: Learning to drive from real-world demonstrations using policy gradients. In Conference on Robot Learning , pages 718–728. PMLR,

  19. [28]

    Safety-enhanced autonomous driving using inter- pretable sensor fusion transformer

    Hao Shao, Letian Wang, Ruobing Chen, Hongsheng Li, and Yu Liu. Safety-enhanced autonomous driving using inter- pretable sensor fusion transformer. In Conference on Robot Learning, pages 726–737. PMLR, 2023. 2

    2023

  20. [29]

    Hpnet: Dynamic trajectory fore- casting with historical prediction attention, 2024

    Xiaolong Tang, Meina Kan, Shiguang Shan, Zhilong Ji, Jin- feng Bai, and Xilin Chen. Hpnet: Dynamic trajectory fore- casting with historical prediction attention, 2024. 2

    2024

  21. [30]

    Con- gested traffic states in empirical observations and micro- scopic simulations

    Martin Treiber, Ansgar Hennecke, and Dirk Helbing. Con- gested traffic states in empirical observations and micro- scopic simulations. Physical review E, 62(2):1805, 2000. 5, 6, 12

    2000

  22. [31]

    Safetynet: Safe planning for real-world self-driving vehicles using machine-learned policies

    Matt Vitelli, Yan Chang, Yawei Ye, Ana Ferreira, Maciej Wołczyk, Bła˙zej Osi´nski, Moritz Niendorf, Hugo Grimmett, Qiangui Huang, Ashesh Jain, et al. Safetynet: Safe planning for real-world self-driving vehicles using machine-learned policies. In 2022 International Conference ...

    2022

  23. [32]

    Improving au- tonomous driving safety with pop: A framework for accurate partially observed trajectory predictions

    Sheng Wang, Yingbing Chen, Jie Cheng, Xiaodong Mei, Ren Xin, Yongkang Song, and Ming Liu. Improving au- tonomous driving safety with pop: A framework for accurate partially observed trajectory predictions. In 2024 IEEE In- ternational Conference on Robotics and Automation (ICR...

    2024

  24. [33]

    Dcms: Motion forecasting with dual consistency and multi-pseudo-target supervision

    Maosheng Ye, Jiamiao Xu, Xunnong Xu, Tengfei Wang, Tongyi Cao, and Qifeng Chen. Dcms: Motion forecasting with dual consistency and multi-pseudo-target supervision. arXiv preprint arXiv:2204.05859, 2022. 2

    2022 arXiv

  25. [34]

    Tnt: Target-driven trajectory pre- diction

    Hang Zhao, Jiyang Gao, Tian Lan, Chen Sun, Ben Sapp, Balakrishnan Varadarajan, Yue Shen, Yi Shen, Yuning Chai, Cordelia Schmid, et al. Tnt: Target-driven trajectory pre- diction. In Conference on Robot Learning, pages 895–904. PMLR, 2021. 2

    2021

  26. [35]

    Query-centric trajectory prediction

    Zikang Zhou, Jianping Wang, Yung-Hui Li, and Yu-Kai Huang. Query-centric trajectory prediction. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, pages 17863–17873, 2023. 2

    2023

  27. [36]

    Experiments on NuPlan a

    Supplementary Material A. Experiments on NuPlan a. Implementation Details Dynamic Encoding. Our model comprehensively encodes the state of each agent within the environment at time t us- ing a vector si t defined as (pi t, θi t, vi t, bi t, Ii t ), where p and θ denote positio...

  28. [37]

    Feature extraction targets map elements and agents within a 120- meter radius, adhering to the NuPlan challenge specifica- tions

    as the foundational backbone of the planner. Feature extraction targets map elements and agents within a 120- meter radius, adhering to the NuPlan challenge specifica- tions. The planning timeframe is set to 8 seconds, with 2 seconds of historical data for optimal efficiency. ...

  29. [38]

    The ego plan is shown in yellow and pink, while agent predictions are depicted in green and blue

    Rectangles represent vehicles, with the orange rectan- gle indicating the autonomous ego vehicle and green rect- angles denoting dynamic agents such as vehicles, cyclists, and pedestrians. The ego plan is shown in yellow and pink, while agent predictions are depicted in green ...

Pith tools

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