Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
175 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
42 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Multi-Predictor Fusion: Combining Learning-based and Rule-based Trajectory Predictors (2307.01408v1)

Published 3 Jul 2023 in cs.RO, cs.LG, cs.SY, and eess.SY

Abstract: Trajectory prediction modules are key enablers for safe and efficient planning of autonomous vehicles (AVs), particularly in highly interactive traffic scenarios. Recently, learning-based trajectory predictors have experienced considerable success in providing state-of-the-art performance due to their ability to learn multimodal behaviors of other agents from data. In this paper, we present an algorithm called multi-predictor fusion (MPF) that augments the performance of learning-based predictors by imbuing them with motion planners that are tasked with satisfying logic-based rules. MPF probabilistically combines learning- and rule-based predictors by mixing trajectories from both standalone predictors in accordance with a belief distribution that reflects the online performance of each predictor. In our results, we show that MPF outperforms the two standalone predictors on various metrics and delivers the most consistent performance.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (33)
  1. Trajectron++: Dynamically-feasible trajectory forecasting with heterogeneous data. In Proc. European Conf. on Computer Vision, pages 683–700. Springer, 2020.
  2. Scept: Scene-consistent, policy-based trajectory predictions for planning. In Proc. IEEE Conf. on Computer Vision and Pattern Recognition, pages 17103–17112, 2022.
  3. Agentformer: Agent-aware transformers for socio-temporal multi-agent forecasting. In Proc. IEEE Int. Conf. on Computer Vision, pages 9813–9823, 2021.
  4. Liability, ethics, and culture-aware behavior specification using rulebooks. In Proc. IEEE Conf. on Robotics and Automation, pages 8536–8542, 2019.
  5. Rule-based optimal control for autonomous driving. In Proc. ACM/IEEE Int. Conf. on Cyber-Physical Systems, pages 143–154, 2021.
  6. Receding horizon planning with rule hierarchies for autonomous vehicles. arXiv preprint arXiv:2212.03323, 2022.
  7. The reasonable crowd: Towards evidence-based and interpretable models of driving behavior. In Proc. IEEE/RSJ Int. Conf. on Intelligent Robots & Systems, pages 6708–6715, 2021.
  8. P. Grünwald and T. van Ommen. Inconsistency of bayesian inference for misspecified linear models, and a proposal for repairing it. Bayesian Analysis, 12(4):1069–1103, 2017.
  9. D. Helbing and P. Molnar. Social force model for pedestrian dynamics. Physical review E, 51(5):4282, 1995.
  10. Vehicle trajectory prediction based on motion model and maneuver recognition. In Proc. IEEE/RSJ Int. Conf. on Intelligent Robots & Systems, pages 4363–4369. IEEE, 2013.
  11. A survey on motion prediction and risk assessment for intelligent vehicles. ROBOMECH journal, 1(1):1–14, 2014.
  12. Social LSTM: Human trajectory prediction in crowded spaces. In Proc. IEEE Conf. on Computer Vision and Pattern Recognition, pages 961–971, 2016.
  13. Predictionnet: Real-time joint probabilistic traffic prediction for planning, control, and simulation. In Proc. IEEE Conf. on Robotics and Automation, pages 8936–8942, 2022.
  14. Multimodal probabilistic model-based planning for human-robot interaction. In Proc. IEEE Conf. on Robotics and Automation, pages 3399–3406, 2018.
  15. Social GAN: Socially acceptable trajectories with generative adversarial networks. In Proc. IEEE Conf. on Computer Vision and Pattern Recognition, pages 2255–2264, 2018.
  16. Scalability in perception for autonomous driving: Waymo open dataset. In Proc. IEEE Conf. on Computer Vision and Pattern Recognition, pages 2446–2454, 2020.
  17. One thousand and one hours: Self-driving motion prediction dataset. In Proc. Conf. on Robot Learning, pages 409–418. PMLR, 2021.
  18. nuscenes: A multimodal dataset for autonomous driving. In Proc. IEEE Conf. on Computer Vision and Pattern Recognition, pages 11621–11631, 2020.
  19. B. Ivanovic. trajdata: A unified interface to many trajectory forecasting datasets, June 2022. URL https://github.com/nvr-avg/trajdata.
  20. Learning to predict vehicle trajectories with model-based planning. In Proc. Conf. on Robot Learning, pages 1035–1045. PMLR, 2022.
  21. Planning-inspired hierarchical trajectory prediction for autonomous driving. arXiv preprint arXiv:2304.11295, 2023.
  22. Differentiable logic layer for rule guided trajectory prediction. In Proc. Conf. on Robot Learning, pages 2178–2194. PMLR, 2021.
  23. H. A. P. Blom. An efficient filter for abruptly changing systems. In Proc. IEEE Conf. on Decision and Control, pages 656–658, 1984.
  24. H. A. P. Blom. An efficient decision-making-free filter for processes with abrupt changes. IFAC Proceedings Volumes, 18(5):631–636, 1985.
  25. N. Schneider and D. M. Gavrila. Pedestrian path prediction with recursive bayesian filters: A comparative study. In Proc. German Conf. on Pattern Recognition, pages 174–183. Springer, 2013.
  26. An interacting multiple model for trajectory prediction of intelligent vehicles in typical road traffic scenario. IEEE Transactions on Neural Networks and Learning Systems, 2021.
  27. O. Maler and D. Nickovic. Monitoring temporal properties of continuous signals. In Y. Lakhnech and S. Yovine, editors, Formal Techniques, Modelling and Analysis of Timed and Fault-Tolerant Systems, pages 152–166. Springer Berlin Heidelberg, 2004.
  28. Back-propagation through signal temporal logic specifications: Infusing logical structure into gradient-based methods. arXiv preprint arXiv:2008.00097, 2020.
  29. Tree-structured policy planning with learned behavior models. arXiv preprint arXiv:2301.11902, 2023.
  30. Diffstack: A differentiable and modular control stack for autonomous vehicles. In Proc. Conf. on Robot Learning, pages 2170–2180. PMLR, 2023.
  31. nuPlan: A closed-loop ml-based planning benchmark for autonomous vehicles. In Proc. IEEE Conf. on Computer Vision and Pattern Recognition ADP3 workshop, 2021.
  32. A. Majumdar and M. Pavone. How should a robot assess risk? towards an axiomatic theory of risk in robotics. In Proc. Int. Symp. on Robotics Research, pages 75–84. Springer, 2020.
  33. Optimization of conditional value-at-risk. Journal of Risk, 2:21–42, 2000.
Citations (3)
View on Semantic Scholar

Summary

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

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