Pith. sign in

REVIEW 3 major objections 5 minor 58 references

Gradient-based Trajectory Optimization with Parallelized Differentiable Traffic Simulation

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read A parallelized differentiable IDM simulator lets gradient-based optimization filter, reconstruct, and predict traffic trajectories at up to two million vehicles in real time, with physically valid motions.

desk verdict Genuinely useful parallel differentiable IDM simulator, but the 'guaranteed physical plausibility' claim rests on a false softplus inequality that needs a quick correction. read the letter →

arxiv 2412.16750 v2 pith:63B7TIUR submitted 2024-12-21 cs.RO cs.DCcs.LG

Sanghyun Son , Laura Zheng , Brian Clipp , Connor Greenwell , Sujin Philip , Ming C. Lin This is my paper · ORCID
classification cs.ROcs.DCcs.LG
keywords differentiablesimulationIntelligentDriverModeltrajectoryoptimizationfilteringreconstructionpredictionparallelcomputingautonomousdriving
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper claims that the Intelligent Driver Model (IDM), a standard car-following ODE, can be turned into a parallelized, differentiable simulation layer that supports gradient-based trajectory optimization at city scale. The authors modify IDM with softplus lower bounds so that optimal spacing stays non-negative and acceleration is clamped to a physically valid interval, guaranteeing that the next speed is never negative under Euler integration. They show this layer can filter noisy trajectories, reconstruct dense trajectories from sparse observations, and produce training-free trajectory forecasts on the Waymo Open Motion Dataset, all with trajectories that respect the acceleration limits. The practical payoff is that planners and learners can optimize traffic behavior with gradients while trusting that the resulting motions remain physically plausible.

What carries the argument

The machine is a differentiable IDM layer: for each vehicle, acceleration is computed by the IDM formula from gap $\Delta p$ and speed difference $\Delta v$, then passed through softplus lower-bound operations—$s^*_\mathrm{opt} = \log(1+\exp(s_\mathrm{opt}))$ and $a^*_i = a_\mathrm{lb} + \log(1+\exp(a_i - a_\mathrm{lb}))$—which keep spacing non-negative and acceleration in $[a_\mathrm{min}, a_\mathrm{max}]$. Euler integration advances position and speed; the per-vehicle independence of these computations is exploited by packaging each vehicle's state as an independent block so the whole simulation runs in parallel on CPU or GPU.

What would settle it

Fix the per-step gap and speed-difference variables to the actual observed leading-vehicle states on a dense NGSIM trajectory, optimize only the five IDM parameters, and compare the fit to the free-variable version; a large degradation in positional error would show that the car-following model itself is not identifiable and that the bounds alone are producing the 'realistic' trajectories. A second check is to run the learned IDM parameters forward on a holdout trajectory and see whether speeds actually stay non-negative and accelerations stay within $[a_\mathrm{min}, a_\mathrm{max}]$ over 80-frame rollouts.

Watch

Extended reading notes

Core claim

The central claim is that a differentiable IDM-based traffic simulator can simulate up to two million vehicles in real time on CPU or GPU and still provide exact gradients through every simulation step. To make IDM differentiable and physically safe, the paper clamps the optimal gap $s_\mathrm{opt}$ below at zero and the acceleration $a_i$ below at $a_\mathrm{lb} = \max(-v_i/\Delta t, a_\mathrm{min})$, implemented with softplus so the gradient flow is smooth, and it proves by the Euler update that speeds stay non-negative. Using this layer, the paper reports 0% physically implausible trajectories in filtering and reconstruction on NGSIM and a sparse Singapore dataset, and a training-free WOMD forecasting baseline with 0.354 mAP from only 1 second of history and five fitted parameters.

Load-bearing premise

The realism guarantee rests on the assumption that bounding acceleration and keeping speeds non-negative is sufficient for physical plausibility, even though the per-step gap and speed-difference variables are free to adjust during optimization and are not tied to a consistent leading vehicle.

Editorial extensions

If this is right

  • Trajectory filtering and reconstruction can be formulated as gradient-based optimization over IDM parameters plus per-step gap and speed-difference variables, producing trajectories with zero acceleration-limit violations.
  • Traffic simulation can scale to 2 million vehicles per timestep in real time on commodity hardware, making city-scale optimization tractable.
  • A training-free trajectory forecasting baseline becomes available: fitting five IDM parameters to 1 second of history and rolling out 8 seconds reaches 0.354 mAP on WOMD, a reference point for learned predictors.
  • The differentiable layer can be composed with neural networks in end-to-end systems, since gradients flow through the ODE solver.

Reading between the lines

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

  • Because the filtering and reconstruction optimization treats $\Delta p_k$ and $\Delta v_k$ as free per-step variables, a stricter test would fix them to the observed leader states and optimize only IDM parameters; if the fit degrades sharply, the per-step variables are carrying the trajectory fit rather than the car-following model.
  • The WOMD prediction results are dominated by straight-ahead motion along lane centers, so the method's true strength is likely in highway-like settings; evaluating on lane-changing and turning scenarios would clarify whether the leader-projection step preserves the interactions IDM needs.
  • The same softplus-bounded IDM layer could serve as a physics-informed inductive bias inside learned trajectory predictors, replacing the post-hoc filtering that many deep models currently apply.
  • A natural extension is to replace the fixed per-step leader variables with a differentiable leader-assignment layer, which would let the model learn when lane changes occur rather than assuming proximity-based leaders.
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

3 major / 5 minor

Summary. This paper presents a parallelized differentiable traffic simulator based on the Intelligent Driver Model (IDM). The authors introduce softplus-based modifications to IDM intended to prevent negative speeds and to keep accelerations within physical bounds, and they demonstrate that the simulator can process up to 2 million vehicles per timestep on CPU or GPU. Using gradient-based optimization, they apply the simulator to trajectory filtering, dense reconstruction from sparse data, and training-free trajectory forecasting, with experiments on NGSIM and the Waymo Open Motion Dataset. The central claims are that the differentiable IDM layer 'guarantees' physically realistic vehicle motion, that this enables large-scale trajectory optimization, and that the resulting training-free forecasting baseline is competitive given that no neural network is used.

Significance. If the claims are fully correct, the paper is a useful systems contribution: the code is public, the parallelization scheme is simple and effective, and the training-free forecasting baseline is an interesting reference point for learning-based methods. The efficiency results and the filtering/reconstruction pipeline are concrete and likely reproducible from the provided artifacts. However, the central mathematical guarantee is overstated, and the evaluation protocol for prediction requires clarification before the reported numbers can be taken at face value. The core idea of a parallel differentiable IDM layer is sound and worth publishing after correction.

major comments (3)
  1. [Section III.C] The sentence 'Note that ai(t)* is guaranteed to exist in [amin, amax]' is mathematically false. Because log(1+exp(x)) > x for every finite x, ai* = alb + log(1+exp(ai - alb)) > ai. In particular, when the raw IDM acceleration ai equals amax, ai* exceeds amax. The lower-bound part of the construction is valid: ai* >= alb, and the Euler update then yields non-negative speed. But the upper bound is not guaranteed. Since the abstract and contributions repeatedly promise 'guaranteed' physical validity, this overstatement is load-bearing. Please either replace the softplus form with a function that is bounded above, e.g., add a hard upper clamp, or explicitly rephrase the guarantee to cover only non-negative speeds and lower-bounded decelerations.
  2. [Section IV.A] The optimization over per-step variables Δp_k and Δv_k, initialized to 10 and 0 and updated freely at every simulation step, effectively reduces the IDM constraint to a bounded-acceleration smoother: the trajectory is no longer forced to respect a consistent car-following relation to a leader, and the five IDM parameters are not identifiable from the reconstruction loss alone. This weakens the claim in Section V.B.6 that the method 'can infer driver behavior during optimization' and the broader claim that the generated motions are realistic in the IDM sense. Please clarify the role of these per-step variables, report an ablation without this per-step freedom, and discuss identifiability or add regularization on Δp_k and Δv_k.
  3. [Section V.C] The prediction evaluation is inconsistent about the agent population. Section IV.B states 'For each vehicle active in traffic lanes, we evaluate their performance,' but Section V.C states 'we consider all benchmark trajectories and agent classes in the validation set, including that of cyclists and pedestrians.' Since the simulator is built on vehicle car-following and lane-center projection, evaluating non-vehicle agents with IDM would bias the comparison against SOTA models that explicitly model all classes, whereas reporting vehicle-only metrics would change the baseline comparison. Please specify exactly which agents are evaluated and provide a per-class or vehicle-only breakdown.
minor comments (5)
  1. [Section III.A] Equation (2) uses the IDM exponent δ, but δ is not listed among the hyperparameters and its value is never defined. Please define it or state explicitly that it is fixed to a standard value.
  2. [Section V.B.1] The reconstruction experiments use a non-public dataset of 21,750 trajectories. Since the code is public, please state in the main text that this dataset is not released, and indicate whether any public substitute or anonymized version can be made available for reproducibility.
  3. [Section IV.B] The statement 'This task is not feasible with existing simulators' is too strong, as parallelized data-driven simulators such as Waymax and GPUDrive already perform large-scale trajectory simulation, even if they are not ODE-based in the same way.
  4. [Section V.A] The 'real-time' claim is reported as a per-timestep cost of less than 30 ms, but the simulation timestep used in the benchmark is not stated. Please specify the timestep so the real-time interpretation is meaningful.
  5. [Table II] The 'Params' column reports 5 for the proposed method. Please clarify whether the trajectory prediction fitting also optimizes per-step Δp and Δv variables; if not, state explicitly that only the five IDM parameters are fitted for forecasting.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation chain: prediction is a genuine out-of-sample IDM rollout, filtering/reconstruction is explicit fitting under constraints, and self-citations are contextual rather than load-bearing.

full rationale

The derivation chain is not circular. The IDM equations come from the external original car-following model [12], and the paper's differentiable layer is an implementation built on standard Euler integration. The non-negative speed guarantee follows algebraically from the definition alb = max(-vi/dt, amin) combined with Eq. (3), not from a fitted or predicted quantity; it is a designed constraint, not a derived prediction. Trajectory filtering and reconstruction are explicitly posed as optimization problems in which IDM parameters and per-step gap/speed variables are optimized to match data, so reporting low error and 0% implausible trajectories reports the outcome of a fit under enforced bounds rather than a prediction forced by a fitted parameter. The trajectory prediction task is the only forward-looking claim: IDM parameters are fitted to 1-second history and then rolled out for 8 seconds without optimizing future states, so the future trajectory is genuinely out-of-sample and not statistically forced. The self-citations ([2], [28], [31]) are contextual mentions of prior differentiable traffic simulation and gradient-based optimization work; they are not load-bearing, no uniqueness theorem is imported from the authors' prior work, and no ansatz is adopted solely through those citations. One mathematical caveat belongs in correctness, not circularity: Section III.C's assertion that ai* is guaranteed to lie in [amin, amax] is false because softplus(x) > x, so the upper bound can be exceeded; the lower-bound side is valid. Likewise, the 'physical plausibility' guarantee is definitional in that the Imp metric uses the same |a| <= 10 bound that the model enforces, but this is constraint satisfaction rather than a circular prediction. Overall, the central claims are self-contained against external benchmarks and do not reduce to their inputs by construction.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The central results rest on IDM as the behavioral model, Euler integration, the softplus clipping scheme, and for prediction the 1D lane projection. The five IDM parameters and the per-step Δp_k, Δv_k are fitted to data. No new physical entities are introduced. The softplus bound axiom is flagged because the paper's claim that the acceleration lies in [amin, amax] is false.

free parameters (3)
  • amax, apref, Tpref, smin, vtarg = Distributions in Fig. 4; e.g., amax around 6.5-10, smin 4.6-5.4, etc.
    Five IDM driver-behavior parameters optimized per trajectory to minimize reconstruction loss (Section IV.A). These are fitted to data, not derived.
  • Δp_k, Δv_k per timestep = not reported, initialized to 10 and 0
    Per-simulation-step leading gap and relative speed, optimized for each trajectory in filtering/reconstruction (Section IV.A). This is a free control variable per timestep, giving the model enough flexibility to fit arbitrary trajectories.
  • amin = -10 m/s^2
    Fixed maximum deceleration bound (Section III.C). A hyperparameter, not fitted but assumed.
assumptions (5)
  • domain assumption Intelligent Driver Model accurately describes car-following behavior
    The entire simulator is built on IDM (Section III.A). If IDM does not capture real driving, the simulated trajectories may be unrealistic despite the acceleration bounds.
  • domain assumption A single leading vehicle in the same lane determines a vehicle's acceleration
    IDM's core premise, stated in Section III.A.
  • domain assumption Euler integration with small Δt is a sufficient approximation
    The paper uses Euler integration (Eq. 3) and states it works 'with a sufficiently small timestep' (Section III.C).
  • ad hoc to paper Softplus lower bounds guarantee physically valid accelerations
    The paper introduces softplus to enforce bounds, but the claim that the result is in [amin, amax] is false; softplus is unbounded above (Section III.C).
  • domain assumption 1D lane-projection preserves leader-follower relationships for prediction
    In Section IV.B, agents are projected onto lane polylines and IDM is applied in 1D; if the projection loses the ordering or leader relations, the rollout is invalid.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Gradient-based Trajectory Optimization with Parallelized Differentiable Traffic Simulation." pith.science (2026). https://pith.science/paper/63B7TIUR

@misc{pith2026241216750,
  author       = {Pith},
  title        = {Pith review of: Gradient-based Trajectory Optimization with Parallelized Differentiable Traffic Simulation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/63B7TIUR}},
  note         = {Machine review of arXiv:2412.16750}
}
read the original abstract

We present a parallelized differentiable traffic simulator based on the Intelligent Driver Model (IDM), a car-following framework that incorporates driver behavior as key variables. Our vehicle simulator efficiently models vehicle motion, generating trajectories that can be supervised to fit real-world data. By leveraging its differentiable nature, IDM parameters are optimized using gradient-based methods. With the capability to simulate up to 2 million vehicles in real time, the system is scalable for large-scale trajectory optimization. We show that we can use the simulator to filter noise in the input trajectories (trajectory filtering), reconstruct dense trajectories from sparse ones (trajectory reconstruction), and predict future trajectories (trajectory prediction), with all generated trajectories adhering to physical laws. We validate our simulator and algorithm on several datasets including NGSIM and Waymo Open Dataset. The code is publicly available at: https://github.com/SonSang/diffidm.

Figures

Figures reproduced from arXiv: 2412.16750 by the authors.

Figure 1
Figure 1. Computational cost of our traffic simulator. Using either (multi-threaded) CPU or GPU, our simulation takes less than 30 milliseconds per timestep to process 2 million vehicles in both (a) forward and (b) backward pass. differentiable with respect to that of the previous one, en￾abling gradient-based optimization techniques that require fewer samples than gradient-free methods. The differen￾tiable programming paradi… view at source ↗
Figure 2
Figure 2. Trajectory Optimization Problems. For the given data points (red flags), we can fit a simulated trajectory to them by optimizing IDM variables. (a) When data points are dense, we can filter physically inaccurate noises in the original trajectory. (b) When they are sparse, we can reconstruct dense trajectories. (c) We can even predict future trajectories based on the IDM variables. context [42]–[47]. Current state-of… view at source ↗
Figure 3
Figure 3. Overall Framework. In our traffic simulator, for each frame, (a) we first collect variables for each vehicle in the scene to use in IDM. These variables include the position (pi) and velocity (vi) of each vehicle, and its relationship to its leading vehicle (ph(i) , vh(i) ). Since we can apply IDM to each vehicle’s variables independently, we can process them in parallel, rather than sequentially. In (b) and (c), th… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Distribution of optimized IDM parameters in trajectory reconstruction task. Note that the distribution of each parameter adheres well to the general, real-world scenarios. which represent driver behavior and remain fixed for each trajectory. These parameters are initia…
Figure 6
Figure 6. Figure 6: Training-free future rollouts on the Waymo Open Motion Dataset (WOMD). We show qualitative results in trajectory forecasting with our differentiable traffic simulator. In the trajectory forecasting task, we infer the traffic state based on road graph and agent observat…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

58 extracted references · 42 canonical work pages

  1. [1]

    Cityflow: A city-scale benchmark for multi-target multi-camera vehicle tracking and re- identification,

    Z. Tang, M. Naphade, M.-Y . Liu, X. Yang, S. Birchfield, S. Wang, R. Kumar, D. Anastasiu, and J.-N. Hwang, “Cityflow: A city-scale benchmark for multi-target multi-camera vehicle tracking and re- identification,” in Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition , 2019, pp. 8797–8806

    work page 2019

  2. [2]

    Differentiable hybrid traffic simulation,

    S. Son, Y .-L. Qiao, J. Sewall, and M. C. Lin, “Differentiable hybrid traffic simulation,” ACM Transactions on Graphics (TOG) , vol. 41, no. 6, pp. 1–10, 2022

    work page 2022

  3. [3]

    An environment for autonomous driving decision- making,

    E. Leurent, “An environment for autonomous driving decision- making,” https://github.com/eleurent/highway-env, 2018

    2018

  4. [4]

    Flow: Architecture and benchmarking for reinforcement learning in traffic control,

    C. Wu, A. Kreidieh, K. Parvate, E. Vinitsky, and A. M. Bayen, “Flow: Architecture and benchmarking for reinforcement learning in traffic control,” arXiv preprint arXiv:1710.05465 , vol. 10, 2017

    arXiv 2017

  5. [5]

    Car-following theory of steady-state traffic flow,

    D. C. Gazis, R. Herman, and R. B. Potts, “Car-following theory of steady-state traffic flow,” Operations research, vol. 7, no. 4, pp. 499– 505, 1959

    work page 1959

  6. [6]

    Nonlinear follow-the- leader models of traffic flow,

    D. C. Gazis, R. Herman, and R. W. Rothery, “Nonlinear follow-the- leader models of traffic flow,” Operations research, vol. 9, no. 4, pp. 545–567, 1961

    work page 1961

  7. [7]

    Nonlinear effects in the dynamics of car following,

    G. F. Newell, “Nonlinear effects in the dynamics of car following,” Operations research, vol. 9, no. 2, pp. 209–229, 1961

    work page 1961

  8. [8]

    Microscopic modeling of traffic flow: investigation of collision free vehicle dynamics,

    S. Krauss, “Microscopic modeling of traffic flow: investigation of collision free vehicle dynamics,” Apr 1998

    work page 1998

Show all 58 references
  1. [9]

    Dynamical model of traffic congestion and numerical simulation,

    M. Bando, K. Hasebe, A. Nakayama, A. Shibata, and Y . Sugiyama, “Dynamical model of traffic congestion and numerical simulation,” Physical review E , vol. 51, no. 2, p. 1035, 1995

    1995

  2. [10]

    A behavioural car-following model for computer simula- tion,

    P. G. Gipps, “A behavioural car-following model for computer simula- tion,” Transportation research part B: methodological , vol. 15, no. 2, pp. 105–111, 1981

    1981

  3. [11]

    Full velocity difference model for a car-following theory,

    R. Jiang, Q. Wu, and Z. Zhu, “Full velocity difference model for a car-following theory,” Physical Review E , vol. 64, no. 1, p. 017101, 2001

    2001

  4. [12]

    Congested traffic states in empirical observations and microscopic simulations,

    M. Treiber, A. Hennecke, and D. Helbing, “Congested traffic states in empirical observations and microscopic simulations,” Physical review E, vol. 62, no. 2, p. 1805, 2000

    2000

  5. [13]

    On kinematic waves ii. a theory of traffic flow on long crowded roads,

    M. J. Lighthill and G. B. Whitham, “On kinematic waves ii. a theory of traffic flow on long crowded roads,”Proceedings of the royal society of london. series a. mathematical and physical sciences , vol. 229, no. 1178, pp. 317–345, 1955

    1955

  6. [14]

    Shock waves on the highway,

    P. I. Richards, “Shock waves on the highway,” Operations research, vol. 4, no. 1, pp. 42–51, 1956

    1956

  7. [15]

    Model of freeway traffic and control,

    H. J. Payne, “Model of freeway traffic and control,” Mathematical Model of Public System , pp. 51–61, 1971

    1971

  8. [16]

    G. B. Whitham, Linear and nonlinear waves . John Wiley & Sons, 2011

    2011

  9. [17]

    Resurrection of

    A. Aw and M. Rascle, “Resurrection of” second order” models of traffic flow,” SIAM journal on applied mathematics , vol. 60, no. 3, pp. 916–938, 2000

    2000

  10. [18]

    A non-equilibrium traffic model devoid of gas-like behavior,

    H. M. Zhang, “A non-equilibrium traffic model devoid of gas-like behavior,” Transportation Research Part B: Methodological , vol. 36, no. 3, pp. 275–290, 2002

    2002

  11. [19]

    Recent development and applications of sumo-simulation of urban mobility,

    D. Krajzewicz, J. Erdmann, M. Behrisch, and L. Bieker, “Recent development and applications of sumo-simulation of urban mobility,” International journal on advances in systems and measurements , vol. 5, no. 3&4, 2012

    2012

  12. [20]

    Carla: An open urban driving simulator,

    A. Dosovitskiy, G. Ros, F. Codevilla, A. Lopez, and V . Koltun, “Carla: An open urban driving simulator,” in Conference on robot learning . PMLR, 2017, pp. 1–16

    2017

  13. [21]

    Waymax: An accelerated, data- driven simulator for large-scale autonomous driving research,

    C. Gulino, J. Fu, W. Luo, G. Tucker, E. Bronstein, Y . Lu, J. Harb, X. Pan, Y . Wang, X. Chen et al. , “Waymax: An accelerated, data- driven simulator for large-scale autonomous driving research,” Ad- vances in Neural Information Processing Systems , vol. 36, 2024

    2024

  14. [22]

    Moss: A large-scale open microscopic traffic simulation system,

    J. Zhang, W. Ao, J. Yan, C. Rong, D. Jin, W. Wu, and Y . Li, “Moss: A large-scale open microscopic traffic simulation system,” arXiv preprint arXiv:2405.12520, 2024

    2024 arXiv

  15. [23]

    End-to-end differentiable physics for learning and control,

    F. de Avila Belbute-Peres, K. Smith, K. Allen, J. Tenenbaum, and J. Z. Kolter, “End-to-end differentiable physics for learning and control,” Advances in neural information processing systems , vol. 31, 2018

    2018

  16. [24]

    A differentiable physics engine for deep learning in robotics,

    J. Degrave, M. Hermans, J. Dambre, and F. Wyffels, “A differentiable physics engine for deep learning in robotics,” Frontiers in neuro- robotics, vol. 13, p. 6, 2019

    2019

  17. [25]

    Efficient differentiable simulation of articulated bodies,

    Y .-L. Qiao, J. Liang, V . Koltun, and M. C. Lin, “Efficient differentiable simulation of articulated bodies,” in International Conference on Machine Learning. PMLR, 2021, pp. 8661–8671

    2021

  18. [26]

    Brax–a differentiable physics engine for large scale rigid body simulation,

    C. D. Freeman, E. Frey, A. Raichuk, S. Girgin, I. Mordatch, and O. Bachem, “Brax–a differentiable physics engine for large scale rigid body simulation,” arXiv preprint arXiv:2106.13281 , 2021

    2021 arXiv

  19. [27]

    Accelerated policy learning with parallel differen- tiable simulation,

    J. Xu, V . Makoviychuk, Y . Narang, F. Ramos, W. Matusik, A. Garg, and M. Macklin, “Accelerated policy learning with parallel differen- tiable simulation,” arXiv preprint arXiv:2204.07137 , 2022

    2022 arXiv

  20. [28]

    Gradient informed proximal policy optimization,

    S. Son, L. Zheng, R. Sullivan, Y .-L. Qiao, and M. Lin, “Gradient informed proximal policy optimization,” Advances in Neural Informa- tion Processing Systems , vol. 36, 2024

    2024

  21. [29]

    Differentiable agent-based simulation for gradient- guided simulation-based optimization,

    P. Andelfinger, “Differentiable agent-based simulation for gradient- guided simulation-based optimization,” in Proceedings of the 2021 ACM SIGSIM Conference on Principles of Advanced Discrete Simu- lation, 2021, pp. 27–38

    2021

  22. [30]

    Imagining the road ahead: Multi-agent trajectory prediction via differentiable simulation,

    A. ´Scibior, V . Lioutas, D. Reda, P. Bateni, and F. Wood, “Imagining the road ahead: Multi-agent trajectory prediction via differentiable simulation,” in 2021 IEEE International Intelligent Transportation Systems Conference (ITSC) , 2021, pp. 720–725

    2021

  23. [31]

    Traffic-aware autonomous driving with differentiable traffic simulation,

    L. Zheng, S. Son, and M. C. Lin, “Traffic-aware autonomous driving with differentiable traffic simulation,” in 2023 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 2023, pp. 3517–3523

    2023

  24. [32]

    Gpudrive: Data-driven, multi-agent driving simulation at 1 million fps,

    S. Kazemkhani, A. Pandya, D. Cornelisse, B. Shacklett, and E. Vinitsky, “Gpudrive: Data-driven, multi-agent driving simulation at 1 million fps,” Aug. 2024. [Online]. Available: http://arxiv.org/abs/ 2408.01584

    2024 arXiv

  25. [33]

    The multi-agent transport simulation matsim

    MATSim Association, “The multi-agent transport simulation matsim.” [Online]. Available: https://matsim.org/

  26. [34]

    Vissim: Multimodal traffic simulation software

    PTV Group, “Vissim: Multimodal traffic simulation software.” [On- line]. Available: https://www.ptvgroup.com/en-us/products/ptv-vissim

  27. [35]

    Modeling lane-changing decisions with mobil,

    M. Treiber and A. Kesting, “Modeling lane-changing decisions with mobil,” in Traffic and Granular Flow ’07. Springer Berlin Heidelberg, 2009, p. 211–221

    2009

  28. [36]

    Large scale interactive motion forecasting for autonomous driving: The waymo open motion dataset,

    S. Ettinger, S. Cheng, B. Caine, C. Liu, H. Zhao, S. Pradhan, Y . Chai, B. Sapp, C. R. Qi, Y . Zhou, Z. Yang, A. Chouard, P. Sun, J. Ngiam, V . Vasudevan, A. McCauley, J. Shlens, and D. Anguelov, “Large scale interactive motion forecasting for autonomous driving: The waymo ope...

    2021

  29. [37]

    A new methodol- ogy for vehicle trajectory reconstruction based on wavelet analysis,

    M. R. Fard, A. S. Mohaymany, and M. Shahri, “A new methodol- ogy for vehicle trajectory reconstruction based on wavelet analysis,” Transportation Research Part C: Emerging Technologies, vol. 74, pp. 150–167, 2017

    2017

  30. [38]

    Estimating individual speed- spacing relationship and assessing ability of newell’s car-following model to reproduce trajectories,

    A. Duret, C. Buisson, and N. Chiabaut, “Estimating individual speed- spacing relationship and assessing ability of newell’s car-following model to reproduce trajectories,” Transportation research record, vol. 2088, no. 1, pp. 188–197, 2008

    2008

  31. [39]

    Validity of trajectory-based cali- bration approach of car-following models in presence of measurement errors,

    S. Ossen and S. P. Hoogendoorn, “Validity of trajectory-based cali- bration approach of car-following models in presence of measurement errors,” Transportation Research Record , vol. 2088, no. 1, pp. 117– 125, 2008

    2008

  32. [40]

    Estimating acceleration and lane-changing dynamics from next generation simulation trajectory data,

    C. Thiemann, M. Treiber, and A. Kesting, “Estimating acceleration and lane-changing dynamics from next generation simulation trajectory data,” Transportation Research Record, vol. 2088, no. 1, pp. 90–101, 2008

    2008

  33. [41]

    Smoothing methods to minimize impact of global positioning system random error on travel distance, speed, and acceleration profile estimates,

    J. Jun, R. Guensler, and J. H. Ogle, “Smoothing methods to minimize impact of global positioning system random error on travel distance, speed, and acceleration profile estimates,” Transportation Research Record, vol. 1972, no. 1, pp. 141–150, 2006

    1972

  34. [42]

    Motionlm: Multi-agent motion forecasting as language modeling,

    A. Seff, B. Cera, D. Chen, M. Ng, A. Zhou, N. Nayakanti, K. S. Refaat, R. Al-Rfou, and B. Sapp, “Motionlm: Multi-agent motion forecasting as language modeling,” 2023

    2023

  35. [43]

    Wayformer: Motion forecasting via simple & efficient atten- tion networks,

    N. Nayakanti, R. Al-Rfou, A. Zhou, K. Goel, K. S. Refaat, and B. Sapp, “Wayformer: Motion forecasting via simple & efficient atten- tion networks,” in 2023 IEEE International Conference on Robotics and Automation (ICRA) , 2023, pp. 2980–2987

    2023

  36. [44]

    Query-centric trajectory prediction,

    Z. Zhou, J. Wang, Y .-H. Li, and Y .-K. Huang, “Query-centric trajectory prediction,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2023

    2023

  37. [45]

    Mtr++: Multi-agent mo- tion prediction with symmetric scene modeling and guided intention querying,

    S. Shi, L. Jiang, D. Dai, and B. Schiele, “Mtr++: Multi-agent mo- tion prediction with symmetric scene modeling and guided intention querying,” arXiv preprint arXiv:2306.17770 , 2023

    2023 arXiv

  38. [46]

    Motion transformer with global intention localization and local movement refinement,

    ——, “Motion transformer with global intention localization and local movement refinement,” Advances in Neural Information Processing Systems, 2022

    2022

  39. [47]

    Multipath++: Efficient information fusion and trajectory aggregation for behavior prediction,

    B. Varadarajan, A. Hefny, A. Srivastava, K. S. Refaat, N. Nayakanti, A. Cornman, K. Chen, B. Douillard, C. P. Lam, D. Anguelov, and B. Sapp, “Multipath++: Efficient information fusion and trajectory aggregation for behavior prediction,” in 2022 International Conference on Robo...

    2022

  40. [48]

    Scalable differentiable physics for learning and control,

    Y .-L. Qiao, J. Liang, V . Koltun, and M. C. Lin, “Scalable differentiable physics for learning and control,” in ICML, 2020

    2020

  41. [49]

    Mtr v3: 1st place solution for 2024 waymo open dataset challenge - motion prediction,

    C. Shi, S. Shi, and L. Jiang, “Mtr v3: 1st place solution for 2024 waymo open dataset challenge - motion prediction,” 2024

    2024

  42. [50]

    Controlmtr: Control-guided motion transformer with scene-compliant intention points for feasible motion prediction,

    J. Sun, C. Yuan, S. Sun, S. Wang, Y . Han, S. Ma, Z. Huang, A. Wong, K. P. Tee, and M. H. Ang Jr, “Controlmtr: Control-guided motion transformer with scene-compliant intention points for feasible motion prediction,” arXiv preprint arXiv:2404.10295 , 2024

    2024 arXiv

  43. [51]

    Scaling mo- tion forecasting models with ensemble distillation,

    S. Ettinger, K. Goel, A. Srivastava, and R. Al-Rfou, “Scaling mo- tion forecasting models with ensemble distillation,” arXiv preprint arXiv:2404.03843, 2024

    2024 arXiv

  44. [52]

    Machine learning for autonomous vehi- cle’s trajectory prediction: A comprehensive survey, challenges, and future research directions,

    V . Bharilya and N. Kumar, “Machine learning for autonomous vehi- cle’s trajectory prediction: A comprehensive survey, challenges, and future research directions,” Vehicular Communications , p. 100733, 2024

    2024

  45. [53]

    Pytorch: An imperative style, high-performance deep learning library,

    A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga et al. , “Pytorch: An imperative style, high-performance deep learning library,” Advances in neural information processing systems , vol. 32, 2019

    2019

  46. [54]

    The next generation simulation program,

    V . Alexiadis, J. Colyar, J. Halkias, R. Hranac, and G. McHale, “The next generation simulation program,” Institute of Transportation Engineers. ITE Journal , vol. 74, no. 8, p. 22, 2004

    2004

  47. [55]

    Scipy 1.0: fundamental algorithms for scientific computing in python,

    P. Virtanen, R. Gommers, T. E. Oliphant, M. Haberland, T. Reddy, D. Cournapeau, E. Burovski, P. Peterson, W. Weckesser, J. Bright et al., “Scipy 1.0: fundamental algorithms for scientific computing in python,” Nature methods, vol. 17, no. 3, pp. 261–272, 2020

    2020

  48. [56]

    Adam: A method for stochastic optimization,

    D. P. Kingma, “Adam: A method for stochastic optimization,” arXiv preprint arXiv:1412.6980, 2014

    2014 arXiv

  49. [57]

    Eda: Evolving and distinct anchors for multimodal motion prediction,

    L. Lin, X. Lin, T. Lin, L. Huang, R. Xiong, and Y . Wang, “Eda: Evolving and distinct anchors for multimodal motion prediction,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 38, no. 4, 2024, pp. 3432–3440

    2024

  50. [58]

    Microscopic traffic simulation using sumo,

    P. A. Lopez, M. Behrisch, L. Bieker-Walz, J. Erdmann, Y .-P. Fl¨otter¨od, R. Hilbrich, L. L ¨ucken, J. Rummel, P. Wagner, and E. Wießner, “Microscopic traffic simulation using sumo,” in The 21st IEEE International Conference on Intelligent Transportation Systems . IEEE, 2018. ...

    2018

Pith tools

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