Dynamically Expanding Capacity of Autonomous Driving with Near-Miss Focused Training Framework
On the map: End-to-end autonomous driving → Adversarial scenarios and safety testing
Abstract (arXiv)
The long-tail distribution of real driving data poses challenges for training and testing autonomous vehicles (AV), where rare yet crucial safety-critical scenarios are infrequent. And virtual simulation offers a low-cost and efficient solution. This paper proposes a near-miss focused training framework for AV. Utilizing the driving scenario information provided by sensors in the simulator, we design novel reward functions, which enable background vehicles (BV) to generate near-miss scenarios and ensure gradients exist not only in collision-free scenes but also in collision scenarios. And then leveraging the Robust Adversarial Reinforcement Learning (RARL) framework for simultaneous training of AV and BV to gradually enhance AV and BV capabilities, as well as generating near-miss scenarios tailored to different levels of AV capabilities. Results from three testing strategies indicate that the proposed method generates scenarios closer to near-miss, thus enhancing the capabilities of both AVs and BVs throughout training.
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Adversarial scenarios and safety testing, one of 9 directions in End-to-end autonomous driving. Builds on 13 of its 18 references that are in the map. First posted on arXiv 5 Jun 2024.
Builds on
13 of 18 references.
Closest ancestors · 5
Same direction, newest first
- 2023Stackelberg Driver Model for Continual Policy Improvement in Scenario-Based Closed-Loop Autonomous Driving25 Sep
- 2023Continual Driving Policy Optimization with Closed-Loop Individualized Curricula25 Sep
- 2023(Re)2H2O: Autonomous Driving Scenario Generation via Reversely Regularized Hybrid Offline-and-Online Reinforcement Learning27 Feb
- 2021DR2L: Surfacing Corner Cases to Robustify Autonomous Driving via Domain Randomization Reinforcement Learning25 Jul
- 2021Corner Case Generation and Analysis for Safety Assessment of Autonomous Vehicles6 Feb
Other references · 8
- 2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actorcited by 4,635
- 2017CARLA: An Open Urban Driving Simulatorcited by 2,431
- 2017Robust Adversarial Reinforcement Learningcited by 439
- 2022A Survey on Safety-Critical Driving Scenario Generation -- A Methodological Perspectivecited by 100
- 2019Testing Scenario Library Generation for Connected and Automated Vehicles, Part I: Methodologycited by 24
- 2020Training Adversarial Agents to Exploit Weaknesses in Deep Control Policiescited by 24
- 2019CMTS: Conditional Multiple Trajectory Synthesizer for Generating Safety-critical Driving Scenarioscited by 19
- 2019Rapidly-exploring Random Trees-based Test Generation for Autonomous Vehiclescited by 1
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2021Adversarial Deep Reinforcement Learning for Improving the Robustness of Multi-agent Autonomous Driving Policies
- 2024Generating Critical Scenarios for Testing Automated Driving Systems
- 2026From Attacks to Curricula: Learnability-Guided Adversarial Training for Safe Autonomous Driving
- 2024Adversarial Safety-Critical Scenario Generation using Naturalistic Human Driving Priors
- 2025Controllable Latent Diffusion for Traffic Simulation
- 2024AdvDiffuser: Generating Adversarial Safety-Critical Driving Scenarios via Guided Diffusion
- 2025Robust Driving Control for Autonomous Vehicles: An Intelligent General-sum Constrained Adversarial Reinforcement Learning Approach
- 2026Learning Responsibility-Attributed Adversarial Scenarios for Testing Autonomous Vehicles