Learning to Drive via Asymmetric Self-Play

• URL: http://arxiv.org/abs/2409.18218v1
• Date: Thu, 26 Sep 2024 18:55:38 GMT
• Title: Learning to Drive via Asymmetric Self-Play
• Authors: Chris Zhang, Sourav Biswas, Kelvin Wong, Kion Fallah, Lunjun Zhang, Dian Chen, Sergio Casas, Raquel Urtasun,
• Abstract summary: We propose asymmetric self-play to scale beyond real data with challenging, solvable, and realistic synthetic scenarios. Our approach pairs a teacher that learns to generate scenarios it can solve but the student cannot, with a student that learns to solve them. Our policies further zero-shot transfer to generate training data for end-to-end autonomy, significantly outperforming state-of-the-art adversarial approaches.
• Score: 34.56873945538085
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large-scale data is crucial for learning realistic and capable driving policies. However, it can be impractical to rely on scaling datasets with real data alone. The majority of driving data is uninteresting, and deliberately collecting new long-tail scenarios is expensive and unsafe. We propose asymmetric self-play to scale beyond real data with additional challenging, solvable, and realistic synthetic scenarios. Our approach pairs a teacher that learns to generate scenarios it can solve but the student cannot, with a student that learns to solve them. When applied to traffic simulation, we learn realistic policies with significantly fewer collisions in both nominal and long-tail scenarios. Our policies further zero-shot transfer to generate training data for end-to-end autonomy, significantly outperforming state-of-the-art adversarial approaches, or using real data alone. For more information, visit https://waabi.ai/selfplay .

Related papers

• ReGentS: Real-World Safety-Critical Driving Scenario Generation Made Stable [88.08120417169971]
  Machine learning based autonomous driving systems often face challenges with safety-critical scenarios that are rare in real-world data. This work explores generating safety-critical driving scenarios by modifying complex real-world regular scenarios through trajectory optimization. Our approach addresses unrealistic diverging trajectories and unavoidable collision scenarios that are not useful for training robust planner.
  arXiv  Detail & Related papers  (2024-09-12T08:26:33Z)
• Imitation Is Not Enough: Robustifying Imitation with Reinforcement Learning for Challenging Driving Scenarios [147.16925581385576]
  We show how imitation learning combined with reinforcement learning can substantially improve the safety and reliability of driving policies. We train a policy on over 100k miles of urban driving data, and measure its effectiveness in test scenarios grouped by different levels of collision likelihood.
  arXiv  Detail & Related papers  (2022-12-21T23:59:33Z)
• Embedding Synthetic Off-Policy Experience for Autonomous Driving via Zero-Shot Curricula [48.58973705935691]
  We show that an agent trained using only a 10% subset of the data performs just as well as an agent trained on the entire dataset. We then demonstrate that this difficulty score can be used in a zero-shot transfer to generate curricula for an imitation-learning based planning agent.
  arXiv  Detail & Related papers  (2022-12-02T18:57:21Z)
• Learning Interactive Driving Policies via Data-driven Simulation [125.97811179463542]
  Data-driven simulators promise high data-efficiency for driving policy learning. Small underlying datasets often lack interesting and challenging edge cases for learning interactive driving. We propose a simulation method that uses in-painted ado vehicles for learning robust driving policies.
  arXiv  Detail & Related papers  (2021-11-23T20:14:02Z)
• Towards Optimal Strategies for Training Self-Driving Perception Models in Simulation [98.51313127382937]
  We focus on the use of labels in the synthetic domain alone. Our approach introduces both a way to learn neural-invariant representations and a theoretically inspired view on how to sample the data from the simulator. We showcase our approach on the bird's-eye-view vehicle segmentation task with multi-sensor data.
  arXiv  Detail & Related papers  (2021-11-15T18:37:43Z)
• Learning to Simulate on Sparse Trajectory Data [26.718807213824853]
  We present a novel framework ImInGAIL to address the problem of learning to simulate the driving behavior from sparse real-world data. To the best of our knowledge, we are the first to tackle the data sparsity issue for behavior learning problems.
  arXiv  Detail & Related papers  (2021-03-22T13:42:11Z)
• TrafficSim: Learning to Simulate Realistic Multi-Agent Behaviors [74.67698916175614]
  We propose TrafficSim, a multi-agent behavior model for realistic traffic simulation. In particular, we leverage an implicit latent variable model to parameterize a joint actor policy. We show TrafficSim generates significantly more realistic and diverse traffic scenarios as compared to a diverse set of baselines.
  arXiv  Detail & Related papers  (2021-01-17T00:29:30Z)
• Enhanced Transfer Learning for Autonomous Driving with Systematic Accident Simulation [3.2456691142503256]
  We show that transfer learning on simulated data sets provide better generalization and collision avoidance. Our results illustrate that information from a model trained on simulated data can be inferred to a model trained on real-world data.
  arXiv  Detail & Related papers  (2020-07-23T17:27:00Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.