Pith. sign in

REVIEW 5 major objections 4 minor 36 references

Active Probing with Multimodal Predictions for Motion Planning

T0 review · 5 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A planner that actively probes other drivers merges successfully 98% of the time, versus 62% for a waiting planner.

desk verdict Active probing is a worthwhile idea, but the printed Wasserstein formula in Lemma 1 is wrong; the paper needs a corrective revision before the empirical claims can be trusted. read the letter →

arxiv 2507.09822 v4 pith:NBQN5BBZ submitted 2025-07-13 cs.RO cs.SYeess.SY

Darshan Gadginmath , Farhad Nawaz , Minjun Sung , Faizan M Tariq , Sangjae Bae , David Isele , Fabio Pasqualetti , Jovin D'sa This is my paper
classification cs.ROcs.SYeess.SY
keywords motionplanningactiveprobingmultimodalpredictionriskmetricWassersteindistancebehaviorinferenceparticlefilterModelPredictiveControl
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

An autonomous planner can deliberately choose gentle accelerating or steering probes to make surrounding drivers reveal whether they are yielding or aggressive, and folding that information into a multimodal risk-aware objective produces safer, faster merging and intersection crossing than passively waiting. The paper builds one objective out of a utility cost, a Wasserstein-based safety risk that has a closed form when prediction modes are Gaussian, and an information-gain term that updates a particle-filter belief over each agent's reward weights through a Boltzmann model that depends on the ego's own actions. In MetaDrive simulations, the active probing variant completes the lane change in 98% of 200 Monte Carlo episodes against 62% for the chance-constrained MPC baseline, and crosses an unsignalized intersection in 96% of episodes against 89%, with zero collisions. The reason to care is that the result recasts cautious driving as an information-gathering problem: the ego does not have to wait for intentions to be revealed if it can nudge them out.

What carries the argument

The load-bearing object is the modified Boltzmann likelihood of Eq. (11), which ties the probability of each predicted mode $k$ for agent $i$ to the reward that mode would give agent $i$ under the ego's candidate action $u_e$. Because the reward model $R^i(x^i,u^i,s^i(x_e,u_e))$ depends on the ego's trajectory through the interaction set $s^i$, the planner's information-gain objective becomes a function of its own controls, so gradient-based MPC can select probing actions. The second essential piece is Lemma 1: with Gaussian mode distributions, the 2-Wasserstein distance in the risk metric is finite and explicit, which keeps the safety cost tractable and lets a risk threshold $\tau$ switch probing off in dangerous modes.

What would settle it

Run the lane-change Monte Carlo with true agents that optimize a reward containing a feature absent from Eq. (15), for example a politeness or gap-acceptance term, and measure whether probing still beats the no-probing variant on success rate and merge time; if the advantage collapses, the probing claim depends on the assumed reward family.

Watch

Extended reading notes

Core claim

The central claim is that active probing with multimodal predictions is both tractable and effective: the ego vehicle can estimate each other agent's behavior parameters $\phi^i$ online and use the estimate to plan. The paper defines a risk $r_i^k(t) = p_i^k(1 + e^{-\alpha W(P^e(t), P_i^k(t))})$ over each predicted mode, with $W$ the 2-Wasserstein distance, and proves (Lemma 1) that when the ego and mode distributions are Gaussian, the risk is always finite and computable in closed form. It then makes the predictor's mode likelihoods action-dependent through a modified Boltzmann model $\hat{p}_i^k(x_e,u_e) \propto p_i^k e^{R^i}$, so that the ego's choice of control changes the expected information gain about $\phi^i$. The experiments support the claim in the paper's own comparison: 98% versus 62% lane-change success, 96% versus 89% intersection success, and faster merge and crossing times than the chance-constrained planner.

Load-bearing premise

The whole probing scheme assumes that other drivers choose actions to maximize the ego's specified reward features, namely velocity matching, safe distance, and lane keeping, so that learning the feature weights tells the ego what they will do next.

Editorial extensions

If this is right

  • Active probing removes the need to wait passively: in the paper's lane-change Monte Carlo, merging success rises from 62% with chance-constrained MPC to 98% with probing.
  • Probing also shortens the maneuver: reported mean time to merge drops from 9.27 seconds for the baseline to 6.87 seconds with probing.
  • The closed-form finite risk metric means the safety objective can be evaluated analytically for Gaussian mixture predictions rather than by sampling trajectories.
  • Because interactions are captured through per-agent reward models instead of a joint dynamics model, adding prediction modes does not blow up the optimization's state space.
  • The framework carries over to agents it did not train on: in the intersection scenario with MetaDrive's built-in controllers, probing still achieves the highest success rate, 96%, and zero collisions.

Reading between the lines

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

  • The lane-change gain may depend on feature alignment: the paper samples true aggressive and defensive controllers from the same reward features the ego assumes, so a test with qualitatively different driver features would show how much of the 98% is probing skill versus model-matching.
  • A natural extension is to make the probe gain depend on prediction confidence: when the predictor's mode likelihoods $p_i^k$ are miscalibrated, the Boltzmann rescaling could amplify noise, so a sensitivity study over miscalibrated likelihoods would sharpen the practical claims.
  • The 2.5-second planning horizon and 0.1-second discretization restrict probes to short nudges; real highway merging often requires seconds-long interactions, so a longer-horizon variant would test whether the mechanism still works when the reveal signal arrives later.
  • The risk threshold $\tau$ currently gates probing per mode; an adaptive threshold that raises when traffic density or prediction covariance grows could make the same idea safer in denser scenes.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 4 minor

Summary. The paper proposes a motion-planning framework that integrates MPC with multimodal trajectory predictions and an active-probing mechanism. The ego agent optimizes a weighted sum of utility, a safety reward based on a Wasserstein-type risk over predicted Gaussian mixture modes, and an information-gain objective that estimates other agents' behavior parameters through a Boltzmann likelihood and a particle filter. The framework is evaluated in MetaDrive in a lane-change scenario and an unsignalized-intersection scenario. The lane-change experiments generate aggressive/defensive agents from a reward model whose feature structure matches the ego's assumed model, while the intersection experiments use MetaDrive's independent RL controllers. Tables II and III report higher success rates and faster merge/cross times for the active-probing planner compared with a chance-constrained MPC baseline and a no-probing variant.

Significance. If the reported results hold, the paper demonstrates a practically relevant benefit of active probing: 98% versus 62% lane-change success and 96% versus 89% intersection success, with zero intersection collisions, alongside released code and videos. The intersection scenario with independent RL controllers is a valuable generality check that goes beyond an in-model-class evaluation. However, the central planner specification currently contains a load-bearing error in the risk formula, and the information-gain update is underspecified, so the experimental numbers cannot yet be attributed to the stated algorithm without correction.

major comments (5)
  1. [III-B, Eq. (8)] The printed closed-form expression cannot be a 2-Wasserstein distance. For equal means and identity covariances, Eq. (8) evaluates to -2 Tr(3I) < 0 instead of 0, whereas the true squared 2-Wasserstein distance is 0. The standard Gaussian formula is W_2^2 = ||\bar{x}_e - \hat{x}_i^k||^2 + Tr(C_e + C_i - 2(C_e^{1/2} C_i C_e^{1/2})^{1/2}). Since Eq. (7) and the safety cost (9) are built on this quantity, the planner whose results appear in Tables II and III is not fully specified. Please correct Eq. (8) and confirm that the experimental implementation uses the corrected expression; the 'always finite' statement in Lemma 1 does not address the sign error.
  2. [III-C, Eqs. (11)-(12) and Algorithm 1] The particle-filter measurement update is under-specified. The posterior b'_k is written only as proportional to b(\phi) \hat{p}^i_k(x_e, u_e), but \hat{p}^i_k depends on \phi through R_i in Eq. (10); the exact likelihood function, its normalization, the resampling step, and the precise KL estimator over M particles are not given. Without these details, the information-gain objective (12) cannot be reproduced from the text, and it is unclear whether the reported probing benefit comes from the stated belief update.
  3. [Table I and Eq. (12)] The quoted risk threshold \tau = 5 can never trigger, because r^i_k(t) = p^i_k(1 + e^{-\alpha W}) \le 2 for all t when p^i_k \le 1 and the exponential term is at most 1. Thus the safety cutoff on probing is inactive in every reported experiment, and any observed benefit of probing cannot be attributed to the threshold mechanism. Please either report the actual threshold used or modify the risk scaling so that the threshold is meaningful.
  4. [IV-C, Table II] The sentence 'our algorithm with probing disabled (\alpha_2 = 0)' is inconsistent with the objective (2), in which \alpha_2 weights the safety term and \alpha_3 weights the information term. If the baseline truly set \alpha_2 = 0, it would have no safety objective; if the intended weight was \alpha_3 = 0, the text should say so. This matters because the 'No probing' row in Table II is one of the two baselines used to attribute the observed improvement to probing.
  5. [IV-A and Eqs. (15)-(16)] The lane-change controllers are generated by optimizing a reward function whose features (velocity matching, inter-agent safety distance, and lane deviation) match the feature structure that the ego assumes in its reward model (15). These experiments therefore primarily validate behavior estimation within the assumed model class. The intersection experiment with MetaDrive RL controllers provides important out-of-model evidence, but the paper should state this limitation explicitly and, if possible, include a lane-change condition with misspecified features.
minor comments (4)
  1. [III-B, Lemma 1] The proof's 'state space is bounded' justification is inconsistent with Gaussian distributions over an unbounded state space; finiteness of the risk follows from finite means and covariances, so the proof should be reworded.
  2. [Table I] The table omits the risk sensitivity \alpha from Eq. (7), the prior covariance \Sigma, and the particle count M used in the belief update; these values are needed to reproduce the experiments.
  3. [Algorithm 1] The threshold check uses r^i_k(t) without a loop over t, so it is unclear whether the cutoff applies if the risk exceeds \tau at any time step, at every time step, or at a particular time; please specify the intended condition.
  4. [Eq. (12)] The time dependence of Info is ambiguous: the information gain is written without a time index, but the cutoff condition depends on r^i_k(t); please clarify how the per-time-step risk feeds into the information objective.

Circularity Check

1 steps flagged · score 4.0 of 10

Lane-change validation is self-consistent by construction because simulated agents are generated from the same linear reward family the ego assumes; the intersection experiments with MetaDrive RL controllers provide independent, non-circular support.

  1. other [Section IV-A and IV-C, Eqs. (15)-(16), Fig. 4, Table II]
    "we assume that the ego agent uses a reward model for agent i inspired from the Intelligent Driver Model [36] that captures key driving behaviors Ri(xi, ui, si) = ... (15) ... The controllers for these behaviors are generated by optimizing the reward function: Ri(xi, ui, si) = ... (16)."

    The lane-change 'true' agents are generated by optimizing (16), a linear reward over velocity-match, inter-vehicle distance, and lane-deviation; the ego's assumed model (15) is the same linear feature family. Thus the ground-truth phi lies inside the estimator's model class, and the Boltzmann/particle-filter posterior (11) is built from that same reward. Fig. 4's correct identification and Table II's probing advantage (98% vs 82%) are therefore partly self-consistency: the estimator recovers a parameter encoded into the simulator under its own likelihood. This is circular validation, not an independent test of probing under model mismatch. The MetaDrive RL intersection experiments (Sec. IV-D) provide external grounding and keep the central claim partly independent.

full rationale

No load-bearing self-citations were found: GLK [35] is used as an input predictor rather than as a justification of the framework, and no uniqueness theorem is imported from prior author work. The central mathematical derivation (risk, closed-form Wasserstein, Boltzmann belief update) is not circular; it is a set of stated modeling choices. The only significant circularity is in the lane-change evaluation, where the data-generating reward (16) lies in the same feature family as the ego's assumed reward (15), making the 'successful probing identifies yielding behavior' result partly true by construction. This is offset by the unsignalized-intersection experiments, where the ego still uses (15) but the true controllers are MetaDrive's RL agents, an external benchmark within the simulator; those results (96% success, 0% collisions in Table III) independently support the framework's utility. Note that Lemma 1's Eq. (8) contains a sign error in the Wasserstein formula (it evaluates negative for identical distributions); that is a correctness concern, not a circularity, so it is not scored here but should be fixed before the risk computation is considered specified.

Assumptions & free parameters 9 free parameters · 5 assumptions · 0 invented entities

The framework uses several hand-chosen parameters (alpha, L, beta, tau, objective weights) and strong domain assumptions about agent rationality and reward structure. The lane-change validation shares the reward model between the true agents and the ego's estimator, which reduces the independence of the evidence.

free parameters (9)
  • alpha = not specified
    Sensitivity in the risk metric (7), chosen by hand.
  • L = 4 m
    Safe distance in the soft barrier (9), Table I.
  • beta = 0.02
    Barrier sharpness in Eq. (9), Table I.
  • tau = 5
    Risk threshold in Eq. (12), Table I; ineffective because the risk metric is bounded above by 2.
  • alpha1 = 0.9
    Utility weight in objective (2), Table I.
  • alpha2 = 0.9
    Safety weight in objective (2), Table I.
  • alpha3 = 0.1
    Information gain weight in objective (2), Table I.
  • Sigma = not specified
    Initial covariance of the belief distribution over behavior parameters.
  • M = not specified
    Number of particles in the particle filter.
assumptions (5)
  • domain assumption Multimodal predictor provides Gaussian modes with covariances and likelihoods (GLK predictor [35]).
    Central input to the framework; if predictions are unreliable the risk and probing objectives degrade.
  • domain assumption State space is bounded in position, velocity, and control.
    Used in Lemma 1 proof to argue Wasserstein distance is finite.
  • domain assumption Agents are rational, non-adversarial reward maximizers with a known reward feature vector J_i.
    Needed for the Boltzmann model (11) and behavior estimation; stated in Section III-C.
  • ad hoc to paper Mode likelihood depends on ego actions through the modified Boltzmann model (11).
    This is the paper's own model of how agent intentions react to probing; not derived from data.
  • ad hoc to paper 2-Wasserstein distance is an appropriate risk measure for planning.
    Remark 1 justifies the choice, but the logistic mapping from W2 to risk is not derived from safety axioms.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Active Probing with Multimodal Predictions for Motion Planning." pith.science (2026). https://pith.science/paper/NBQN5BBZ

@misc{pith2026250709822,
  author       = {Pith},
  title        = {Pith review of: Active Probing with Multimodal Predictions for Motion Planning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NBQN5BBZ}},
  note         = {Machine review of arXiv:2507.09822}
}
read the original abstract

Navigation in dynamic environments requires autonomous systems to reason about uncertainties in the behavior of other agents. In this paper, we introduce a unified framework that combines trajectory planning with multimodal predictions and active probing to enhance decision-making under uncertainty. We develop a novel risk metric that seamlessly integrates multimodal prediction uncertainties through mixture models. When these uncertainties follow a Gaussian mixture distribution, we prove that our risk metric admits a closed-form solution, and is always finite, thus ensuring analytical tractability. To reduce prediction ambiguity, we incorporate an active probing mechanism that strategically selects actions to improve its estimates of behavioral parameters of other agents, while simultaneously handling multimodal uncertainties. We extensively evaluate our framework in autonomous navigation scenarios using the MetaDrive simulation environment. Results demonstrate that our active probing approach successfully navigates complex traffic scenarios with uncertain predictions. Additionally, our framework shows robust performance across diverse traffic agent behavior models, indicating its broad applicability to real-world autonomous navigation challenges. Code and videos are available at https://darshangm.github.io/papers/active-probing-multimodal-predictions/.

Figures

Figures reproduced from arXiv: 2507.09822 by the authors.

Figure 1
Figure 1. Illustrative figure of a merging scenario with uncertain [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Risk assessment with multiple agent trajectories. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Lane change scenarios demonstrating our algorithm’s adaptation to different vehicle behaviors. The ego vehicle (green [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Particle filter estimation of safety weights over time. [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Lane change experiments with time-varying agent [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Unsignalized intersection scene with N = 7 vehicles. Ego vehicle’s perspective (left), and its corresponding bird’s eye view (right). V. CONCLUSION We present a trajectory planning algorithm that combines multimodal predictions with active probing to estimate and adapt…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

36 extracted references · 31 canonical work pages

  1. [1]

    F. M. Tariq, D. Isele, J. S. Baras, and S. Bae. Slas: Speed and lane advisory system for highway navigation. In 2022 IEEE 61st Conference on Decision and Control (CDC) , pages 6979–6986. IEEE, 2022

    work page 2022

  2. [2]

    S. Shi, L. Jiang, D. Dai, and B. Schiele. Mtr++: Multi-agent motion prediction with symmetric scene modeling and guided intention query- ing. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024

    work page 2024

  3. [3]

    Nayakanti, R

    N. Nayakanti, R. Al-Rfou, A. Zhou, K. Goel, K. S. Refaat, and B. Sapp. Wayformer: Motion forecasting via simple & efficient at- tention networks. In 2023 IEEE International Conference on Robotics and Automation (ICRA) , pages 2980–2987, 2023

    work page 2023

  4. [4]

    Zhang, X

    K. Zhang, X. Feng, L. Wu, and Z. He. Trajectory prediction for autonomous driving using spatial-temporal graph attention trans- former. IEEE Transactions on Intelligent Transportation Systems , 23(11):22343–22353, 2022

    work page 2022

  5. [5]

    Salzmann, B

    T. Salzmann, B. Ivanovic, P. Chakravarty, and M. Pavone. Tra- jectron++: Dynamically-feasible trajectory forecasting with hetero- geneous data. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XVIII 16, pages 683–700. Springer, 2020

    work page 2020

  6. [6]

    C. Liu, S. He, H. Liu, and J. Chen. Intention-aware denoising diffusion model for trajectory prediction. arXiv preprint arXiv:2403.09190 , 2024

    arXiv 2024

  7. [7]

    Z. Li, H. Liang, H. Wang, X. Zheng, J. Wang, and P. Zhou. A multi- modal vehicle trajectory prediction framework via conditional diffu- sion model: A coarse-to-fine approach. Knowledge-Based Systems , 280:110990, 2023

    work page 2023

  8. [8]

    L. Feng, M. Bahari, K. M. B. Amor, ´E. Zablocki, M. Cord, and A. Alahi. Unitraj: A unified framework for scalable vehicle trajectory prediction. arXiv preprint arXiv:2403.15098 , 2024

    arXiv 2024

Show all 36 references
  1. [9]

    J. Li, J. Li, S. Bae, and D. Isele. Adaptive prediction ensemble: Improving out-of-distribution generalization of motion forecasting. IEEE Robotics and Automation Letters , 2024

    2024

  2. [10]

    Q. Li, Z. Peng, Z. Xue, Q. Zhang, and B. Zhou. Metadrive: Composing diverse driving scenarios for generalizable reinforcement learning. arXiv preprint arXiv:2109.12674 , 2021

    2021 arXiv

  3. [11]

    Zhang, Y

    Z. Zhang, Y . Zeng, W. Jiang, Y . Pan, and J. Tang. Intention recognition for multiple agents. Information Sciences, 628:360–376, 2023

    2023

  4. [12]

    A. Cui, S. Casas, A. Sadat, R. Liao, and R. Urtasun. Lookout: Diverse multi-future prediction and planning for self-driving. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 16107–16116, 2021

    2021

  5. [13]

    H. Gong, J. Sim, M. Likhachev, and J. Shi. Multi-hypothesis motion planning for visual object tracking. In 2011 International Conference on Computer Vision , pages 619–626. IEEE, 2011

    2011

  6. [14]

    H. Bai, S. Cai, N. Ye, D. Hsu, and W. S. Lee. Intention-aware online pomdp planning for autonomous driving in a crowd. In 2015 ieee international conference on robotics and automation (icra), pages 454–

    2015

  7. [15]

    Qi and S.-C

    S. Qi and S.-C. Zhu. Intent-aware multi-agent reinforcement learning. In 2018 IEEE international conference on robotics and automation (ICRA), pages 7533–7540. IEEE, 2018

    2018

  8. [16]

    R. Tian, L. Sun, M. Tomizuka, and D. Isele. Anytime game-theoretic planning with active reasoning about humans’ latent states for human- centered robots. In 2021 IEEE International Conference on Robotics and Automation (ICRA) , pages 4509–4515. IEEE, 2021

    2021

  9. [17]

    Y . Chen, U. Rosolia, W. Ubellacker, N. Csomay-Shanklin, and A. D. Ames. Interactive multi-modal motion planning with branch model predictive control. IEEE Robotics and Automation Letters , 7(2):5365– 5372, 2022

    2022

  10. [18]

    Isele, A

    D. Isele, A. M. Anon, F. M. Tariq, G. Yeh, A. Singh, and S. Bae. Delayed-decision motion planning in the presence of multiple predic- tions. In IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2025

    2025

  11. [19]

    K. Ren, H. Ahn, and M. Kamgarpour. Chance-constrained trajectory planning with multimodal environmental uncertainty. IEEE Control Systems Letters, 7:13–18, 2022

    2022

  12. [20]

    J. P. Alsterda, M. Brown, and J. C. Gerdes. Contingency model predictive control for automated vehicles. In 2019 American control conference (ACC), pages 717–722. IEEE, 2019

    2019

  13. [21]

    S. H. Nair, H. Lee, E. Joa, Y . Wang, H. E. Tseng, and F. Borrelli. Predictive control for autonomous driving with uncertain, multimodal predictions. IEEE Transactions on Control Systems Technology, 2024

    2024

  14. [22]

    H. Ahn, C. Chen, I. M. Mitchell, and M. Kamgarpour. Safe mo- tion planning against multimodal distributions based on a scenario approach. IEEE Control Systems Letters , 6:1142–1147, 2021

    2021

  15. [23]

    Laine, D

    F. Laine, D. Fridovich-Keil, C.-Y . Chiu, and C. Tomlin. Multi- hypothesis interactions in game-theoretic motion planning. In 2021 IEEE International Conference on Robotics and Automation (ICRA) , pages 8016–8023. IEEE, 2021

    2021

  16. [24]

    A. Mesbah. Stochastic model predictive control with active uncertainty learning: A survey on dual control. Annual Reviews in Control , 45:107–117, 2018

    2018

  17. [25]

    H. Hu, D. Isele, S. Bae, and J. F. Fisac. Active uncertainty reduction for safe and efficient interaction planning: A shielding-aware dual control approach. The International Journal of Robotics Research , 43(9):1382–1408, 2024

    2024

  18. [26]

    Knaup, J

    J. Knaup, J. D’sa, B. Chalaki, T. Naes, H. N. Mahjoub, E. Moradi- Pari, and P. Tsiotras. Active learning with dual model predictive path- integral control for interaction-aware autonomous highway on-ramp merging. In 2024 IEEE International Conference on Robotics and Automation...

    2024

  19. [27]

    S. Wang, Y . Lyu, and J. M. Dolan. Active probing and influencing human behaviors via autonomous agents. In 2023 IEEE International Conference on Robotics and Automation (ICRA) , pages 1514–1521. IEEE, 2023

    2023

  20. [28]

    Peyr ´e, M

    G. Peyr ´e, M. Cuturi, et al. Computational optimal transport: With applications to data science. Foundations and Trends® in Machine Learning, 11(5-6):355–607, 2019

    2019

  21. [29]

    Majumdar and M

    A. Majumdar and M. Pavone. How should a robot assess risk? towards an axiomatic theory of risk in robotics. In Robotics Research: The 18th International Symposium ISRR , pages 75–84. Springer, 2019

    2019

  22. [30]

    F. M. Tariq, D. Isele, J. S. Baras, and S. Bae. Rcms: Risk-aware crash mitigation system for autonomous vehicles. In 2023 IEEE 26th International Conference on Intelligent Transportation Systems (ITSC), pages 3950–3957. IEEE, 2023

    2023

  23. [31]

    Baimukashev, G

    D. Baimukashev, G. Alcan, and V . Kyrki. Automated feature selection for inverse reinforcement learning. arXiv preprint arXiv:2403.15079 , 2024

    2024 arXiv

  24. [32]

    B. D. Ziebart, A. L. Maas, J. A. Bagnell, A. K. Dey, et al. Maximum entropy inverse reinforcement learning. In Aaai, volume 8, pages 1433–1438. Chicago, IL, USA, 2008

    2008

  25. [33]

    J. F. Fisac, A. Bajcsy, S. L. Herbert, D. Fridovich-Keil, S. Wang, C. J. Tomlin, and A. D. Dragan. Probabilistically safe robot planning with confidence-based human predictions. arXiv preprint arXiv:1806.00109, 2018

    2018 arXiv

  26. [34]

    Frank, J

    M. Frank, J. Leitner, M. Stollenga, A. F ¨orster, and J. Schmidhuber. Curiosity driven reinforcement learning for motion planning on hu- manoids. Frontiers in neurorobotics, 7:25, 2014

    2014

  27. [35]

    Isele, P

    D. Isele, P. Gupta, X. Liu, and S. Bae. Gaussian lane keeping: A robust prediction baseline. arXiv preprint arXiv:2407.18451 , 2024

    2024 arXiv

  28. [36]

    Treiber, A

    M. Treiber, A. Hennecke, and D. Helbing. Congested traffic states in empirical observations and microscopic simulations. Physical review E, 62(2):1805, 2000

    2000

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.