Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
110 tokens/sec
GPT-4o
56 tokens/sec
Gemini 2.5 Pro Pro
44 tokens/sec
o3 Pro
6 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

CaDRE: Controllable and Diverse Generation of Safety-Critical Driving Scenarios using Real-World Trajectories (2403.13208v2)

Published 19 Mar 2024 in cs.RO

Abstract: Simulation is an indispensable tool in the development and testing of autonomous vehicles (AVs), offering an efficient and safe alternative to road testing. An outstanding challenge with simulation-based testing is the generation of safety-critical scenarios, which are essential to ensure that AVs can handle rare but potentially fatal situations. This paper addresses this challenge by introducing a novel framework, CaDRE, to generate realistic, diverse, and controllable safety-critical scenarios. Our approach optimizes for both the quality and diversity of scenarios by employing a unique formulation and algorithm that integrates real-world scenarios, domain knowledge, and black-box optimization. We validate the effectiveness of our framework through extensive testing in three representative types of traffic scenarios. The results demonstrate superior performance in generating diverse and high-quality scenarios with greater sample efficiency than existing reinforcement learning (RL) and sampling-based methods.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (55)
  1. 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,” Advances in Neural Information Processing Systems, vol. 36, 2024.
  2. J. Herman, J. Francis, S. Ganju, B. Chen, A. Koul, A. Gupta, A. Skabelkin, I. Zhukov, M. Kumskoy, and E. Nyberg, “Learn-to-race: A multimodal control environment for autonomous racing,” in proceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 9793–9802.
  3. S. H. Park, G. Lee, J. Seo, M. Bhat, M. Kang, J. Francis, A. Jadhav, P. P. Liang, and L.-P. Morency, “Diverse and admissible trajectory forecasting through multimodal context understanding,” in Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XI 16.   Springer, 2020, pp. 282–298.
  4. M. Xu, Z. Liu, P. Huang, W. Ding, Z. Cen, B. Li, and D. Zhao, “Trustworthy reinforcement learning against intrinsic vulnerabilities: Robustness, safety, and generalizability,” arXiv preprint arXiv:2209.08025, 2022.
  5. P. Huang, X. Zhang, Z. Cao, S. Liu, M. Xu, W. Ding, J. Francis, B. Chen, and D. Zhao, “What went wrong? closing the sim-to-real gap via differentiable causal discovery,” in Conference on Robot Learning.   PMLR, 2023, pp. 734–760.
  6. W. Ding, C. Xu, M. Arief, H. Lin, B. Li, and D. Zhao, “A survey on safety-critical driving scenario generation—a methodological perspective,” IEEE Transactions on Intelligent Transportation Systems, 2023.
  7. B. Stoler, I. Navarro, M. Jana, S. Hwang, J. Francis, and J. Oh, “Safeshift: Safety-informed distribution shifts for robust trajectory prediction in autonomous driving,” arXiv preprint arXiv:2309.08889, 2023.
  8. E. Bronstein, S. Srinivasan, S. Paul, A. Sinha, M. O’Kelly, P. Nikdel, and S. Whiteson, “Embedding synthetic off-policy experience for autonomous driving via zero-shot curricula,” in Conference on Robot Learning.   PMLR, 2023, pp. 188–198.
  9. P. Huang, M. Xu, F. Fang, and D. Zhao, “Robust reinforcement learning as a stackelberg game via adaptively-regularized adversarial training,” arXiv preprint arXiv:2202.09514, 2022.
  10. M. Xu, P. Huang, Y. Niu, V. Kumar, J. Qiu, C. Fang, K.-H. Lee, X. Qi, H. Lam, B. Li et al., “Group distributionally robust reinforcement learning with hierarchical latent variables,” in International Conference on Artificial Intelligence and Statistics.   PMLR, 2023, pp. 2677–2703.
  11. L. Feng, Q. Li, Z. Peng, S. Tan, and B. Zhou, “Trafficgen: Learning to generate diverse and realistic traffic scenarios,” in 2023 IEEE International Conference on Robotics and Automation (ICRA).   IEEE, 2023, pp. 3567–3575.
  12. S. Tan, K. Wong, S. Wang, S. Manivasagam, M. Ren, and R. Urtasun, “Scenegen: Learning to generate realistic traffic scenes,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 892–901.
  13. P. Huang, M. Xu, J. Zhu, L. Shi, F. Fang, and D. Zhao, “Curriculum reinforcement learning using optimal transport via gradual domain adaptation,” Advances in Neural Information Processing Systems, vol. 35, pp. 10 656–10 670, 2022.
  14. J. Wang, A. Pun, J. Tu, S. Manivasagam, A. Sadat, S. Casas, M. Ren, and R. Urtasun, “Advsim: Generating safety-critical scenarios for self-driving vehicles,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 9909–9918.
  15. N. Hanselmann, K. Renz, K. Chitta, A. Bhattacharyya, and A. Geiger, “King: Generating safety-critical driving scenarios for robust imitation via kinematics gradients,” in European Conference on Computer Vision.   Springer, 2022, pp. 335–352.
  16. Z. Zhong, D. Rempe, D. Xu, Y. Chen, S. Veer, T. Che, B. Ray, and M. Pavone, “Guided conditional diffusion for controllable traffic simulation,” in 2023 IEEE International Conference on Robotics and Automation (ICRA).   IEEE, 2023, pp. 3560–3566.
  17. S. Tan, B. Ivanovic, X. Weng, M. Pavone, and P. Kraehenbuehl, “Language conditioned traffic generation,” arXiv preprint arXiv:2307.07947, 2023.
  18. M. Xu, P. Huang, F. Li, J. Zhu, X. Qi, K. Oguchi, Z. Huang, H. Lam, and D. Zhao, “Scalable safety-critical policy evaluation with accelerated rare event sampling,” in 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).   IEEE, 2022, pp. 12 919–12 926.
  19. W. Ding, B. Chen, M. Xu, and D. Zhao, “Learning to collide: An adaptive safety-critical scenarios generating method,” in 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).   IEEE, 2020, pp. 2243–2250.
  20. W. Ding, H. Lin, B. Li, K. J. Eun, and D. Zhao, “Semantically adversarial driving scenario generation with explicit knowledge integration,” arXiv preprint arXiv:2106.04066, vol. 1, 2021.
  21. M. Xu, P. Huang, W. Yu, S. Liu, X. Zhang, Y. Niu, T. Zhang, F. Xia, J. Tan, and D. Zhao, “Creative robot tool use with large language models,” arXiv preprint arXiv:2310.13065, 2023.
  22. S. Suo, S. Regalado, S. Casas, and R. Urtasun, “Trafficsim: Learning to simulate realistic multi-agent behaviors,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 10 400–10 409.
  23. Q. Li, Z. Peng, L. Feng, C. Duan, W. Mo, B. Zhou et al., “Scenarionet: Open-source platform for large-scale traffic scenario simulation and modeling,” arXiv preprint arXiv:2306.12241, 2023.
  24. W. Ding, M. Xu, and D. Zhao, “Cmts: Conditional multiple trajectory synthesizer for generating safety-critical driving scenarios,” in International Conference on Robotics and Automation (ICRA).   IEEE, 2020.
  25. W. Ding, W. Wang, and D. Zhao, “Multi-vehicle trajectories generation for vehicle-to-vehicle encounters,” in 2019 IEEE International Conference on Robotics and Automation (ICRA), 2019.
  26. C. Xu, W. Ding, W. Lyu, Z. Liu, S. Wang, Y. He, H. Hu, D. Zhao, and B. Li, “Safebench: A benchmarking platform for safety evaluation of autonomous vehicles,” Advances in Neural Information Processing Systems, vol. 35, pp. 25 667–25 682, 2022.
  27. W. Ding, B. Chen, B. Li, K. J. Eun, and D. Zhao, “Multimodal safety-critical scenarios generation for decision-making algorithms evaluation,” IEEE Robotics and Automation Letters, vol. 6, no. 2, pp. 1551–1558, 2021.
  28. W. Ding, H. Lin, B. Li, and D. Zhao, “Causalaf: Causal autoregressive flow for goal-directed safety-critical scenes generation,” arXiv preprint arXiv:2110.13939, 2021.
  29. M. Klischat and M. Althoff, “Generating critical test scenarios for automated vehicles with evolutionary algorithms,” in 2019 IEEE Intelligent Vehicles Symposium (IV).   IEEE, 2019, pp. 2352–2358.
  30. M. Arief, Z. Huang, G. K. S. Kumar, Y. Bai, S. He, W. Ding, H. Lam, and D. Zhao, “Deep probabilistic accelerated evaluation: A robust certifiable rare-event simulation methodology for black-box safety-critical systems,” in International Conference on Artificial Intelligence and Statistics.   PMLR, 2021, pp. 595–603.
  31. Y. Cao, C. Xiao, A. Anandkumar, D. Xu, and M. Pavone, “Advdo: Realistic adversarial attacks for trajectory prediction,” in European Conference on Computer Vision.   Springer, 2022, pp. 36–52.
  32. C. Zhang, J. Tu, L. Zhang, K. Wong, S. Suo, and R. Urtasun, “Learning realistic traffic agents in closed-loop,” in 7th Annual Conference on Robot Learning, 2023.
  33. W. Ding, Y. Cao, D. Zhao, C. Xiao, and M. Pavone, “Realgen: Retrieval augmented generation for controllable traffic scenarios,” arXiv preprint arXiv:2312.13303, 2023.
  34. W. Ding, H. Lin, B. Li, and D. Zhao, “Generalizing goal-conditioned reinforcement learning with variational causal reasoning,” Advances in Neural Information Processing Systems, vol. 35, pp. 26 532–26 548, 2022.
  35. A. Li, S. Chen, L. Sun, N. Zheng, M. Tomizuka, and W. Zhan, “Scegene: Bio-inspired traffic scenario generation for autonomous driving testing,” IEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 9, pp. 14 859–14 874, 2021.
  36. J. Achiam, S. Adler, S. Agarwal, L. Ahmad, I. Akkaya, F. L. Aleman, D. Almeida, J. Altenschmidt, S. Altman, S. Anadkat et al., “Gpt-4 technical report,” arXiv preprint arXiv:2303.08774, 2023.
  37. Z. Zhong, D. Rempe, Y. Chen, B. Ivanovic, Y. Cao, D. Xu, M. Pavone, and B. Ray, “Language-guided traffic simulation via scene-level diffusion,” arXiv preprint arXiv:2306.06344, 2023.
  38. J.-B. Mouret and J. Clune, “Illuminating search spaces by mapping elites,” arXiv preprint arXiv:1504.04909, 2015.
  39. B. Tjanaka, M. C. Fontaine, D. H. Lee, Y. Zhang, N. R. Balam, N. Dennler, S. S. Garlanka, N. D. Klapsis, and S. Nikolaidis, “Pyribs: A bare-bones python library for quality diversity optimization,” in Proceedings of the Genetic and Evolutionary Computation Conference, ser. GECCO ’23.   New York, NY, USA: Association for Computing Machinery, 2023, p. 220–229. [Online]. Available: https://doi.org/10.1145/3583131.3590374
  40. A. Cully, J. Clune, D. Tarapore, and J.-B. Mouret, “Robots that can adapt like animals,” Nature, vol. 521, no. 7553, pp. 503–507, 2015.
  41. V. Bhatt, H. Nemlekar, M. Fontaine, B. Tjanaka, H. Zhang, Y.-C. Hsu, and S. Nikolaidis, “Surrogate assisted generation of human-robot interaction scenarios,” arXiv preprint arXiv:2304.13787, 2023.
  42. Y.-S. Tung, M. B. Luebbers, A. Roncone, and B. Hayes, “Workspace optimization techniques to improve prediction of human motion during human-robot collaboration,” arXiv preprint arXiv:2401.12965, 2024.
  43. A. Morel, Y. Kunimoto, A. Coninx, and S. Doncieux, “Automatic acquisition of a repertoire of diverse grasping trajectories through behavior shaping and novelty search,” in 2022 International Conference on Robotics and Automation (ICRA).   IEEE, 2022, pp. 755–761.
  44. J. Huber, F. Hélénon, M. Kappel, E. Chelly, M. Khoramshahi, F. B. Amar, and S. Doncieux, “Speeding up 6-dof grasp sampling with quality-diversity,” arXiv preprint arXiv:2403.06173, 2024.
  45. S. Surana, B. Lim, and A. Cully, “Efficient learning of locomotion skills through the discovery of diverse environmental trajectory generator priors,” in 2023 IEEE International Conference on Robotics and Automation (ICRA).   IEEE, 2023, pp. 12 134–12 141.
  46. J. Nordmoen, F. Veenstra, K. O. Ellefsen, and K. Glette, “Map-elites enables powerful stepping stones and diversity for modular robotics,” Frontiers in Robotics and AI, vol. 8, p. 639173, 2021.
  47. E. Zardini, D. Zappetti, D. Zambrano, G. Iacca, and D. Floreano, “Seeking quality diversity in evolutionary co-design of morphology and control of soft tensegrity modular robots,” in Genetic and Evolutionary Computation Conference, 2021.
  48. P. Liu, Z. Guo, H. Yu, H. Linghu, Y. Li, Y. Hou, H. Ge, and Q. Zhang, “A preliminary study of multi-task MAP-elites with knowledge transfer for robotic arm design,” in 2022 IEEE Congress on Evolutionary Computation (CEC).   IEEE, jul 2022. [Online]. Available: https://doi.org/10.1109%2Fcec55065.2022.9870374
  49. M. C. Fontaine, J. Togelius, S. Nikolaidis, and A. K. Hoover, “Covariance matrix adaptation for the rapid illumination of behavior space,” in Proceedings of the 2020 genetic and evolutionary computation conference, 2020, pp. 94–102.
  50. H. Caesar, J. Kabzan, K. S. Tan, W. K. Fong, E. Wolff, A. Lang, L. Fletcher, O. Beijbom, and S. Omari, “nuplan: A closed-loop ml-based planning benchmark for autonomous vehicles,” arXiv preprint arXiv:2106.11810, 2021.
  51. M. Fontaine and S. Nikolaidis, “Differentiable quality diversity,” Advances in Neural Information Processing Systems, vol. 34, pp. 10 040–10 052, 2021.
  52. ——, “A quality diversity approach to automatically generating human-robot interaction scenarios in shared autonomy,” arXiv preprint arXiv:2012.04283, 2020.
  53. N. Hansen, “The cma evolution strategy: A tutorial,” arXiv preprint arXiv:1604.00772, 2016.
  54. R. J. Williams, “Simple statistical gradient-following algorithms for connectionist reinforcement learning,” Machine learning, vol. 8, pp. 229–256, 1992.
  55. Y. Liu, P. Ramachandran, Q. Liu, and J. Peng, “Stein variational policy gradient,” arXiv preprint arXiv:1704.02399, 2017.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (6)
  1. Peide Huang (15 papers)
  2. Wenhao Ding (43 papers)
  3. Jonathan Francis (48 papers)
  4. Bingqing Chen (17 papers)
  5. Ding Zhao (172 papers)
  6. Benjamin Stoler (6 papers)
Citations (2)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets