Pith. sign in

REVIEW 4 major objections 6 minor 47 references

Learning Implicit Social Navigation Behavior using Deep Inverse Reinforcement Learning

T0 review · 4 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read Smooth Maximum Entropy Deep Inverse Reinforcement Learning lets a robot learn to yield to oncoming traffic in a narrow crossing from a handful of expert demonstrations, recovering from most unseen states.

desk verdict A practical smoothing-plus-extrapolation trick on top of MEDIRL that looks qualitatively right, but the headline 92% claim needs more statistical support before it is citable as a measured result. read the letter →

arxiv 2501.06946 v1 pith:NK5UABNX submitted 2025-01-12 cs.RO

Tribhi Kathuria , Ke Liu , Junwoo Jang , X. Jessie Yang , Maani Ghaffari This is my paper · ORCID
classification cs.RO
keywords inversereinforcementlearningsocialnavigationrewardfew-shotdemonstrationdeadlockavoidancecrowdnarrowcrossingmobilerobot
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 tries to show that a robot can acquire socially appropriate navigation behavior from a small set of expert demonstrations instead of handcrafted rules or rewards. The proposed method, Smooth Maximum Entropy Deep Inverse Reinforcement Learning (S-MEDIRL), learns a time-varying reward map from top-down scene images, the past trajectories of the robot and human, human heading and velocity, and the robot goal. The reward map is used to sample a reference trajectory that a local collision-avoidance controller then executes. In a photo-realistic narrow-crossing simulation, the learned agent completes about 92 percent of unseen test episodes, while a standard ORCA (Optimal Reciprocal Collision Avoidance) agent deadlocks and a rule-based backoff agent succeeds more slowly. If the result holds, social norms such as yielding and deadlock avoidance could be learned from data rather than specified by a designer.

What carries the argument

The central object is the S-MEDIRL reward predictor, a U-Net that maps a stacked feature image—RGB top-down view, robot and human past trajectories, human heading and velocity, and robot goal—to a reward value for every grid cell at the current time. Training augments the MEDIRL demonstration-matching gradient in two ways: it adds extrapolated trajectories sampled near the demonstrated starts so the network sees more states, and it applies a bilateral filtering loss that penalizes steep reward differences between neighboring cells, acting like a costmap inflation layer. At deployment the predicted reward map is used to sample a reference trajectory that an ORCA controller tracks, replanning every 0.2 seconds.

What would settle it

Train S-MEDIRL on a second scene with different geometry, such as a wider doorway, a corridor, or a T-junction, and test with a human who occasionally pauses, changes speed, or backs off. If the success rate falls well below the reported 92 percent or the robot repeatedly deadlocks, the claimed few-shot generalization and implicit yielding are limited to the demonstrated scene rather than a general social-navigation behavior.

Watch

Extended reading notes

Core claim

On its own terms, the central claim is that S-MEDIRL implicitly learns social behavior: from expert demonstrations of a robot and a human crossing through a narrow doorway, the agent recovers a reward map that encodes both scene navigability and the interaction dynamics, and this map makes the robot yield to oncoming traffic, wait in a sensible place, and avoid deadlock. The reported result is an average success rate of about 92 percent across 13 unseen episodes in the same scene, where nominal ORCA deadlocks in every negotiating episode and the ORCA-backoff agent succeeds but takes roughly thirty seconds longer. The behavior is not produced by an explicit yielding rule; it emerges from the learned reward.

Load-bearing premise

The load-bearing premise is that fifteen expert demonstrations in one simulated narrow-crossing scene, with a human who walks straight and never backs off, are enough to encode the social norms for that scenario, and that thirteen unseen episodes run five times each give a reliable measure of success.

Editorial extensions

If this is right

  • S-MEDIRL recovers from most unseen states and raises the average success rate to about 92 percent on 13 unseen episodes in the narrow-crossing scene.
  • The learned agent yields to oncoming traffic and avoids deadlock without any explicit deadlock detector or handcrafted yielding rule.
  • The agent completes negotiating episodes faster than the ORCA-backoff baseline because it waits at a learned position instead of retreating to its start.
  • The pipeline removes handcrafted reward and feature design for this social behavior, encoding the norm in the learned reward map instead.
  • The reference-trajectory controller still handles local collision avoidance, suggesting the learned reward supplies the global social reasoning that the local controller lacks.

Reading between the lines

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

  • The smoothing loss is functionally a learned inflation layer, so the same device could replace hand-tuned costmap inflation in other learned navigation pipelines.
  • The evaluation covers one human who never concedes; whether the learned yielding transfers to two-way traffic, groups, or people who sometimes yield themselves is a natural next experiment.
  • Because the reward is conditioned on image-like feature channels, retraining for other social scenarios may only require collecting demonstrations in the same top-down format, not redesigning the network.
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 / 6 minor

Summary. This paper proposes Smooth Maximum Entropy Deep Inverse Reinforcement Learning (S-MEDIRL), which extends MEDIRL with demonstration-data extrapolation and a bilateral-filtering smoothing loss on the learned reward map. Training uses top-down image features, robot and human trajectory history, human heading and velocity, and the robot goal from 15 expert demonstrations in a single Habitat-Sim narrow-crossing scene. The learned reward is replanned every 0.2 s and executed by an ORCA local controller. The paper reports that S-MEDIRL achieves about 92% success across 13 unseen episodes versus about 82% for the MEDIRL and ORCA-backoff baselines, and it qualitatively shows yielding and deadlock avoidance.

Significance. If the quantitative claims are supported, the contribution is a useful pipeline for learning socially normative local navigation behavior from few demonstrations without handcrafted reward features. The public release of code and data and the use of a photorealistic simulator are strengths, and the qualitative yielding and deadlock-avoidance behavior is a valuable proof of concept. However, the headline improvement rests on a small, clustered sample with no uncertainty quantification, and the core smoothing-loss contribution lacks a direct ablation, so the significance is not yet established.

major comments (4)
  1. [Section IV-B, Fig. 6] The central success-rate claim is not statistically supported. The evaluation is 13 episodes times 5 runs; no confidence intervals, per-episode counts, or significance tests are given. Under a naive two-proportion analysis, 60/65 versus 53/65 has a standard error near 5.8 percentage points and a two-sided p-value near 0.06, and clustering by episode makes the effective sample smaller. The text also uses "median success rate" for MEDIRL and "average success rate" for S-MEDIRL, which is ambiguous. Please provide the per-episode outcome table, a clearly defined aggregate metric, and uncertainty quantification.
  2. [Section III-C, Fig. 4] No ablation isolates the smoothing loss. S-MEDIRL adds both extrapolation and Eq. (5) relative to MEDIRL, so the 92% result and Fig. 4 do not show that the smoothing loss is beneficial. Add an extrapolation-only condition and a smoothing-loss condition with sensitivity to sigma_s, sigma_c, and the loss weight, using the same success and completion-time metrics.
  3. [Section IV and Section VI] The generalization claim "can recover from most unseen states" is too broad for the evidence: one fixed scene, one scripted human policy that never backs off, and 13 start and goal variations. The social behavior conclusion is therefore limited to the tested narrow-crossing setup. Either add experiments with different scenes and human behaviors or restrict the claims explicitly.
  4. [Section III-C and III-D] The few-shot claim depends on an ambiguous data description. "15 demonstration data points" is not defined; the extrapolation procedure says "six random start positions close to the current start" but gives no number of extrapolated trajectories or frames. Please specify the exact number of expert trajectories, samples per trajectory, extrapolation count, and the split used for training and testing, as this is central to the few-shot claim.
minor comments (6)
  1. [Section IV-A.2] The deadlock-detection sentence is garbled: "deadlock is detected : if two agents move less than 0.1 m over 50 timesteps ( 10 seconded." Please rephrase and correct the typo.
  2. [Section IV-B] The ORCA-backoff agent is described as "trained to choose the robot's start position"; since it is a rule-based baseline, use "configured" or "designed" instead.
  3. [Fig. 6 caption] The caption labels both baselines as "ORCA"; distinguish the nominal ORCA agent from the ORCA-backoff agent in the legend and caption.
  4. [Eq. (5)] Specify the loss weight with which L_bilateral is added to Eq. (3), and give the values of sigma_s, sigma_c, and the neighborhood size used in the experiments.
  5. [Fig. 7] Medians are plotted without error bars or per-episode markers; show the full distribution or at least interquartile ranges so the reader can assess variability.
  6. [Section III-A] The MDP has no discount factor or horizon; clarify how the finite look-ahead of 10 steps is represented in the MDP.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the learned reward and the 92% success claim are measured on held-out episodes, and the self-citations provide only architectural and prior-work context.

full rationale

The paper's derivation chain is not circular. The reward is recovered by MEDIRL-style maximum-entropy gradient matching (Eq. 4) from expert demonstrations plus extrapolated data; the smoothing term (Eq. 5) is a regularizer on the training reward, not a fitted predictor of the test metric. The headline result ("average success rate to about 92%", Sec. IV-B, Fig. 6) is an empirical outcome over 13 unseen episodes with 5 runs each, not a parameter fitted to those episodes; no equation in the paper makes the reported success rate equal to an input by construction. Self-citations appear ([11], [14], [45]), but they are used for U-Net architecture, prior social-navigation context, and the nav-mesh baseline planner; none is invoked as a uniqueness theorem or as the justification for the claimed improvement, and the S-MEDIRL vs MEDIRL/ORCA comparison is evaluated against external baselines. The Discussion honestly notes reliance on good demonstrations and the difficulty of quantifying social quality, and the small test sample and missing error bars are a statistical/generalization limitation, not a circularity. Under the stated rules, no circular step is established.

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

The smoothing loss introduces two hand-picked scales (sigma_s and sigma_c) and the data augmentation relies on a fixed number of random starts and a counter-stop threshold. These parameters are not fitted to the test data, but no sensitivity analysis is given. The core assumptions are the optimality of the demonstrations and the sufficiency and observability of the grid-based state.

free parameters (4)
  • sigma_s (bilateral smoothing spatial scale)
    Controls how strongly spatially distant pixels are pulled together in the smoothing loss (Eq. 5). Chosen by hand; no sensitivity analysis is provided.
  • sigma_c (bilateral smoothing intensity scale)
    Controls how strongly reward values that differ are pulled together in Eq. 5. Chosen by hand; no sensitivity analysis is provided.
  • number of extrapolated start positions = 6
    Six random start positions near the demonstrated start are added to the dataset to augment few-shot demonstrations; this is an ad hoc choice.
  • counter-stop threshold = 3 seconds
    A 3-second threshold is used to identify when the robot is stationary (counter-stop) and to add stopping data; a hand-picked design decision.
assumptions (4)
  • domain assumption Expert demonstrations are near-optimal for the desired social behavior
    The IRL framework assumes demonstrations are drawn from an optimal policy under the true reward. The paper states in Discussion that it "relies heavily on there being good demonstrations available."
  • domain assumption The state representation is fully observable and sufficient
    The MDP uses a 60x60 top-down image with perfect knowledge of robot and human positions; this ignores partial observability and perception noise.
  • domain assumption Deterministic transitions
    Section III-A states the transition between states is assumed deterministic, which may not hold on real hardware with control noise.
  • standard math Maximum entropy IRL framework is valid
    The gradient update in Eq. 4 follows MaxEnt IRL (Ziebart et al.), a standard method.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Learning Implicit Social Navigation Behavior using Deep Inverse Reinforcement Learning." pith.science (2026). https://pith.science/paper/NK5UABNX

@misc{pith2026250106946,
  author       = {Pith},
  title        = {Pith review of: Learning Implicit Social Navigation Behavior using Deep Inverse Reinforcement Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NK5UABNX}},
  note         = {Machine review of arXiv:2501.06946}
}
read the original abstract

This paper reports on learning a reward map for social navigation in dynamic environments where the robot can reason about its path at any time, given agents' trajectories and scene geometry. Humans navigating in dense and dynamic indoor environments often work with several implied social rules. A rule-based approach fails to model all possible interactions between humans, robots, and scenes. We propose a novel Smooth Maximum Entropy Deep Inverse Reinforcement Learning (S-MEDIRL) algorithm that can extrapolate beyond expert demos to better encode scene navigability from few-shot demonstrations. The agent learns to predict the cost maps reasoning on trajectory data and scene geometry. The agent samples a trajectory that is then executed using a local crowd navigation controller. We present results in a photo-realistic simulation environment, with a robot and a human navigating a narrow crossing scenario. The robot implicitly learns to exhibit social behaviors such as yielding to oncoming traffic and avoiding deadlocks. We compare the proposed approach to the popular model-based crowd navigation algorithm ORCA and a rule-based agent that exhibits yielding.

Figures

Figures reproduced from arXiv: 2501.06946 by the authors.

Figure 1
Figure 1. Robots navigating in dense indoor environments with humans in the scene [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Fig. 2a shows the training architecture. The expert demos are sampled every 0.1s to create a new demonstration data point. The past trajectories of the human and robot are fed in to give the network history of past positions. The U-Net architecture feeds into the final regression layer, which outputs the learned reward. The reward is then used to sample a trajectory (Eµ), which is compared to the demonstrated trajec… view at source ↗
Figure 3
Figure 3. The figure shows the augmented noise data fed into the network for one of our [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Comparison of a reward at time t trained with and without the smoothing loss. Fig. 4a shows that there are fewer pixels allocated starkly different values to its neighbors compared to Fig. 4b (a) R Channel (b) G Channel (c) B Channel (d) Robot past (e) Human past (f) H…
Figure 5
Figure 5. Figure 5: This figure provides a snapshot of the network’s input features. The first three [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: The figure shows the success rate of the MEDIRL agents compared to our model-based (ORCA) and rule-based (ORCA) baselines collected over 5 runs. For episodes [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 7
Figure 7. Figure 7: The figure depicts the overall time taken by the robot and the time taken to reach [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]
Figure 8
Figure 8. Figure 8: This figure shows snippets of the robot trajectory at discrete time instants. The [PITH_FULL_IMAGE:figures/full_fig_p007_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

47 extracted references · 35 canonical work pages

  1. [1]

    Learning transferable visual models from natural language supervision,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clark, G. Krueger, and I. Sutskever, “Learning transferable visual models from natural language supervision,” in International Conference on Machine Learning, 2021. [Online]. Available: https://api.semanticscholar.org/ CorpusID:231591445

    work page 2021

  2. [2]

    High-resolution image synthesis with latent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High-resolution image synthesis with latent diffusion models,” Proc. IEEE Conf. Comput. Vis. Pattern Recog. , pp. 10 674–10 685,

    work page

  3. [3]

    A survey of embodied ai: From simulators to research tasks,

    J. Duan, S. Yu, H. L. Tan, H. Zhu, and C. Tan, “A survey of embodied ai: From simulators to research tasks,” IEEE Transactions on Emerging Topics in Computational Intelligence, vol. 6, no. 2, pp. 230–244, 2022

    2022

  4. [4]

    Habitat: A platform for embodied ai research,

    M. Savva, A. Kadian, O. Maksymets, Y . Zhao, E. Wijmans, B. Jain, J. Straub, J. Liu, V . Koltun, J. Malik, et al. , “Habitat: A platform for embodied ai research,” in Proc. IEEE Int. Conf. Comput. Vis. , 2019, pp. 9339–9347

    work page 2019

  5. [5]

    Reinforcement learning in robotics: A survey,

    J. Kober, J. A. Bagnell, and J. Peters, “Reinforcement learning in robotics: A survey,” Int. J. Robot. Res. , vol. 32, no. 11, pp. 1238– 1274, 2013

    work page 2013

  6. [6]

    Gridtopix: Training embodied agents with minimal super- vision,

    U. Jain, I.-J. Liu, S. Lazebnik, A. Kembhavi, L. Weihs, and A. G. Schwing, “Gridtopix: Training embodied agents with minimal super- vision,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) , October 2021, pp. 15 141–15 151

    work page 2021

  7. [7]

    Decentralized non- communicating multiagent collision avoidance with deep reinforce- ment learning,

    Y . F. Chen, M. Liu, M. Everett, and J. P. How, “Decentralized non- communicating multiagent collision avoidance with deep reinforce- ment learning,” in Proc. IEEE Int. Conf. Robot. and Automation . IEEE, 2017, pp. 285–292

    work page 2017

  8. [8]

    An inverse reinforcement learning approach for customizing automated lane change systems,

    J. Liu, L. N. Boyle, and A. G. Banerjee, “An inverse reinforcement learning approach for customizing automated lane change systems,” IEEE Transactions on V ehicular Technology, vol. 71, no. 9, pp. 9261– 9271, 2022

    work page 2022

Show all 47 references
  1. [9]

    Intelligent cooperative collision avoidance at overtaking and lane changing maneuver in 6g- v2x communications,

    S. B. Prathiba, G. Raja, and N. Kumar, “Intelligent cooperative collision avoidance at overtaking and lane changing maneuver in 6g- v2x communications,” IEEE Transactions on V ehicular Technology , vol. 71, no. 1, pp. 112–122, 2021

    2021

  2. [10]

    Feedback-efficient active prefer- ence learning for socially aware robot navigation,

    R. Wang, W. Wang, and B.-C. Min, “Feedback-efficient active prefer- ence learning for socially aware robot navigation,” in 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2022, pp. 11 336–11 343

    2022

  3. [11]

    Solo t-dirl: Socially-aware dynamic local planner based on trajectory-ranked deep inverse reinforcement learning,

    Y . Xu, T. Chakhachiro, T. Kathuria, and M. Ghaffari, “Solo t-dirl: Socially-aware dynamic local planner based on trajectory-ranked deep inverse reinforcement learning,” in 2023 IEEE International Confer- ence on Robotics and Automation (ICRA) . IEEE, 2023, pp. 12 045– 12 051

    2023

  4. [12]

    Socially adaptive path planning in human envi- ronments using inverse reinforcement learning,

    B. Kim and J. Pineau, “Socially adaptive path planning in human envi- ronments using inverse reinforcement learning,” International Journal of Social Robotics , vol. 8, pp. 51–66, 2016

    2016

  5. [13]

    Active preference-based gaussian process regression for reward learning,

    E. Bıyık, N. Huynh, M. J. Kochenderfer, and D. Sadigh, “Active preference-based gaussian process regression for reward learning,” arXiv preprint arXiv:2005.02575 , 2020

    2005 arXiv

  6. [14]

    Energy-based legged robots terrain traversability modeling via deep inverse rein- forcement learning,

    L. Gan, J. W. Grizzle, R. M. Eustice, and M. Ghaffari, “Energy-based legged robots terrain traversability modeling via deep inverse rein- forcement learning,” IEEE Robotics and Automation Letters , vol. 7, no. 4, pp. 8807–8814, 2022

    2022

  7. [15]

    Extrapolating beyond suboptimal demonstrations via inverse reinforcement learning from observations,

    D. S. Brown, W. Goo, P. Nagarajan, and S. Niekum, “Extrapolating beyond suboptimal demonstrations via inverse reinforcement learning from observations,” 2019. [Online]. Available: https://arxiv.org/abs/ 1904.06387

    2019 arXiv

  8. [16]

    Layered costmaps for context-sensitive navigation,

    D. V . Lu, D. Hershberger, and W. D. Smart, “Layered costmaps for context-sensitive navigation,” in 2014 IEEE/RSJ International Conference on Intelligent Robots and Systems . IEEE, 2014, pp. 709– 715

    2014

  9. [17]

    Habitat 3.0: A co-habitat for humans, avatars and robots,

    X. Puig, E. Undersander, A. Szot, M. D. Cote, T.-Y . Yang, R. Partsey, R. Desai, A. W. Clegg, M. Hlavac, S. Y . Min, et al. , “Habitat 3.0: A co-habitat for humans, avatars and robots,” arXiv preprint arXiv:2310.13724, 2023

    2023 arXiv

  10. [18]

    Ros: an open-source robot operating system,

    M. Quigley, K. Conley, B. Gerkey, J. Faust, T. Foote, J. Leibs, R. Wheeler, and A. Y . Ng, “Ros: an open-source robot operating system,” in ICRA workshop on open source software , vol. 3, no. 3.2. Kobe, Japan, 2009, p. 5

    2009

  11. [19]

    Social force model for pedestrian dynam- ics,

    D. Helbing and P. Molnar, “Social force model for pedestrian dynam- ics,” Physical Review E , vol. 51, 05 1998

    1998

  12. [20]

    The walking behaviour of pedestrian social groups and its impact on crowd dynamics,

    M. Moussa ¨ıd, N. Perozo, S. Garnier, D. Helbing, and G. Theraulaz, “The walking behaviour of pedestrian social groups and its impact on crowd dynamics,” PloS one , vol. 5, no. 4, p. e10047, 2010

    2010

  13. [21]

    Walking ahead: The headed social force model,

    F. Farina, D. Fontanelli, A. Garulli, A. Giannitrapani, and D. Prat- tichizzo, “Walking ahead: The headed social force model,” PloS one , vol. 12, no. 1, p. e0169734, 2017

    2017

  14. [22]

    Deep social force,

    S. Kreiss, “Deep social force,” 2021

    2021

  15. [23]

    Reciprocal velocity obstacles for real-time multi-agent navigation,

    J. Van den Berg, M. Lin, and D. Manocha, “Reciprocal velocity obstacles for real-time multi-agent navigation,” in Proc. IEEE Int. Conf. Robot. and Automation . Ieee, 2008, pp. 1928–1935

    2008

  16. [24]

    The dynamic window approach to collision avoidance,

    D. Fox, W. Burgard, and S. Thrun, “The dynamic window approach to collision avoidance,” IEEE Robot. Autom. Mag. , vol. 4, pp. 23–33, 1997

    1997

  17. [25]

    Decentralized social navigation with non-cooperative robots via bi- level optimization,

    R. Chandra, R. Menon, Z. Sprague, A. Anantula, and J. Biswas, “Decentralized social navigation with non-cooperative robots via bi- level optimization,” arXiv preprint arXiv:2306.08815 , 2023

    2023 arXiv

  18. [26]

    After you: doorway negotiation for human-robot and robot-robot interaction,

    J. Thomas and R. Vaughan, “After you: doorway negotiation for human-robot and robot-robot interaction,” in 2018 IEEE/RSJ Interna- tional Conference on Intelligent Robots and Systems (IROS) . IEEE, 2018, pp. 3387–3394

    2018

  19. [27]

    Ds-mpepc: Safe and deadlock- avoiding robot navigation in cluttered dynamic scenes,

    S. H. Arul, J. J. Park, and D. Manocha, “Ds-mpepc: Safe and deadlock- avoiding robot navigation in cluttered dynamic scenes,” arXiv preprint arXiv:2303.10133, 2023

    2023 arXiv

  20. [28]

    Core challenges of social robot navigation: A survey,

    C. Mavrogiannis, F. Baldini, A. Wang, D. Zhao, P. Trautman, A. Stein- feld, and J. Oh, “Core challenges of social robot navigation: A survey,” ACM Transactions on Human-Robot Interaction , vol. 12, no. 3, pp. 1– 39, 2023

    2023

  21. [29]

    Reciprocal n-body collision avoidance,

    J. Van Den Berg, S. J. Guy, M. Lin, and D. Manocha, “Reciprocal n-body collision avoidance,” in Robotics Research: The 14th Interna- tional Symposium ISRR . Springer, 2011, pp. 3–19

    2011

  22. [30]

    Crowd-robot interaction: Crowd-aware robot navigation with attention-based deep reinforce- ment learning,

    C. Chen, Y . Liu, S. Kreiss, and A. Alahi, “Crowd-robot interaction: Crowd-aware robot navigation with attention-based deep reinforce- ment learning,” in 2019 international conference on robotics and automation (ICRA) . IEEE, 2019, pp. 6015–6022

    2019

  23. [31]

    Dipcan: Distilling privi- leged information for crowd-aware navigation

    G. Monaci, M. Aractingi, and T. Silander, “Dipcan: Distilling privi- leged information for crowd-aware navigation.” in Robotics: Science and Systems , 2022

    2022

  24. [32]

    Socially compliant robot navigation in crowded environment by human behavior resemblance using deep reinforcement learning,

    S. S. Samsani and M. S. Muhammad, “Socially compliant robot navigation in crowded environment by human behavior resemblance using deep reinforcement learning,” IEEE Robotics and Automation Letters, vol. 6, no. 3, pp. 5223–5230, 2021

    2021

  25. [33]

    Maximum entropy inverse reinforcement learning

    B. D. Ziebart, A. L. Maas, J. A. Bagnell, A. K. Dey, et al., “Maximum entropy inverse reinforcement learning.” in Proc. AAAI Nat. Conf. Artif. Intell. , vol. 8. Chicago, IL, USA, 2008, pp. 1433–1438

    2008

  26. [34]

    Generative adversarial imitation learning,

    J. Ho and S. Ermon, “Generative adversarial imitation learning,”

  27. [35]

    Guided cost learning: Deep inverse optimal control via policy optimization,

    C. Finn, S. Levine, and P. Abbeel, “Guided cost learning: Deep inverse optimal control via policy optimization,” in International conference on machine learning . PMLR, 2016, pp. 49–58

    2016

  28. [36]

    Learning robust rewards with adversarial inverse reinforcement learning,

    J. Fu, K. Luo, and S. Levine, “Learning robust rewards with adversarial inverse reinforcement learning,” 2018. [Online]. Available: https://arxiv.org/abs/1710.11248

    2018 arXiv

  29. [37]

    Bayesian inverse reinforcement learning

    D. Ramachandran and E. Amir, “Bayesian inverse reinforcement learning.” in IJCAI, vol. 7, 2007, pp. 2586–2591

    2007

  30. [38]

    Maximum entropy deep inverse reinforcement learning,

    M. Wulfmeier, P. Ondruska, and I. Posner, “Maximum entropy deep inverse reinforcement learning,” arXiv preprint arXiv:1507.04888 , 2015

    2015 arXiv

  31. [39]

    Spatiotemporal costmap inference for mpc via deep inverse reinforcement learning,

    K. Lee, D. Isele, E. A. Theodorou, and S. Bae, “Spatiotemporal costmap inference for mpc via deep inverse reinforcement learning,” IEEE Robotics and Automation Letters , vol. 7, no. 2, pp. 3194–3201, 2022

    2022

  32. [40]

    Advanced planning for autonomous vehicles using reinforcement learning and deep inverse reinforcement learning,

    C. You, J. Lu, D. Filev, and P. Tsiotras, “Advanced planning for autonomous vehicles using reinforcement learning and deep inverse reinforcement learning,” Robotics and Autonomous Systems , vol. 114, pp. 1–18, 2019

    2019

  33. [41]

    Medirl: Predicting the visual attention of drivers via maximum entropy deep inverse reinforcement learning,

    S. Baee, E. Pakdamanian, I. Kim, L. Feng, V . Ordonez, and L. Barnes, “Medirl: Predicting the visual attention of drivers via maximum entropy deep inverse reinforcement learning,” in Proceedings of the IEEE/CVF international conference on computer vision , 2021, pp. 13 178–13 188

    2021

  34. [42]

    A deep inverse reinforcement learning ap- proach to route choice modeling with context-dependent rewards,

    Z. Zhao and Y . Liang, “A deep inverse reinforcement learning ap- proach to route choice modeling with context-dependent rewards,” Transportation Research Part C: Emerging Technologies , vol. 149, p. 104079, 2023

    2023

  35. [43]

    Trajectory forecasts in unknown environments conditioned on grid-based plans,

    N. Deo and M. M. Trivedi, “Trajectory forecasts in unknown environments conditioned on grid-based plans,” arXiv preprint arXiv:2001.00735, 2020

    2001 arXiv

  36. [44]

    Learning socially appropriate robot approaching behavior toward groups using deep reinforcement learning,

    Y . Gao, F. Yang, M. Frisk, D. Hemandez, C. Peters, and G. Castellano, “Learning socially appropriate robot approaching behavior toward groups using deep reinforcement learning,” in 2019 28th IEEE inter- national conference on robot and human interactive communication (RO-MAN)...

    2019

  37. [45]

    Providers-clients-robots: Framework for spatial-semantic planning for shared understanding in human-robot interaction,

    T. Kathuria, Y . Xu, T. Chakhachiro, X. J. Yang, and M. Ghaffari, “Providers-clients-robots: Framework for spatial-semantic planning for shared understanding in human-robot interaction,” in 2022 31st IEEE International Conference on Robot and Human Interactive Communication (R...

    2022

  38. [2016]

    Available: https://arxiv.org/abs/1606.03476

    [Online]. Available: https://arxiv.org/abs/1606.03476

    arXiv

  39. [2021]

    Available: https://api.semanticscholar.org/CorpusID: 245335280

    [Online]. Available: https://api.semanticscholar.org/CorpusID: 245335280

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.