mlsanityData to 7 Oct 2026

Diverse and Adaptive Behavior Curriculum for Autonomous Driving: A Student-Teacher Framework with Multi-Agent RL

arXiv 2507.19146 ↗First posted 25 Jul 2025cs.RO cs.LGcited by 0

On the map: End-to-end autonomous driving → Reinforcement learning for driving

Abstract (arXiv)

Autonomous driving faces challenges in navigating complex real-world traffic, requiring safe handling of both common and critical scenarios. Reinforcement learning (RL), a prominent method in end-to-end driving, enables agents to learn through trial and error in simulation. However, RL training often relies on rule-based traffic scenarios, limiting generalization. Additionally, current scenario generation methods focus heavily on critical scenarios, neglecting a balance with routine driving behaviors. Curriculum learning, which progressively trains agents on increasingly complex tasks, is a promising approach to improving the robustness and coverage of RL driving policies. However, existing research mainly emphasizes manually designed curricula, focusing on scenery and actor placement rather than traffic behavior dynamics. This work introduces a novel student-teacher framework for automatic curriculum learning. The teacher, a graph-based multi-agent RL component, adaptively generates traffic behaviors across diverse difficulty levels. An adaptive mechanism adjusts task difficulty based on student performance, ensuring exposure to behaviors ranging from common to critical. The student, though exchangeable, is realized as a deep RL agent with partial observability, reflecting real-world perception constraints. Results demonstrate the teacher's ability to generate diverse traffic behaviors. The student, trained with automatic curricula, outperformed agents trained on rule-based traffic, achieving higher rewards and exhibiting balanced, assertive driving.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Reinforcement learning for driving, one of 9 directions in End-to-end autonomous driving. Builds on 28 of its 35 references that are in the map. First posted on arXiv 25 Jul 2025.

Builds on

28 of 35 references.

Closest ancestors · 3

Same direction, newest first

  1. 2024Autonomous Driving at Unsignalized Intersections: A Review of Decision-Making Challenges and Reinforcement Learning-Based Solutions20 Sep
  2. 2024A Review of Reward Functions for Reinforcement Learning in the context of Autonomous Driving12 Apr
  3. 2022State Dropout-Based Curriculum Reinforcement Learning for Self-Driving at Unsignalized Intersections10 Jul

Other references · 25

  1. 2017Proximal Policy Optimization Algorithmscited by 14,441
  2. 2016Overcoming catastrophic forgetting in neural networkscited by 4,718
  3. 2014Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translationcited by 4,016
  4. 2017CARLA: An Open Urban Driving Simulatorcited by 2,431
  5. 2022Planning-oriented Autonomous Drivingcited by 1,068
  6. 2021The Surprising Effectiveness of PPO in Cooperative, Multi-Agent Gamescited by 903
  7. 2021NuPlan: A closed-loop ML-based planning benchmark for autonomous vehiclescited by 485
  8. 2023End-to-end Autonomous Driving: Challenges and Frontierscited by 485
  9. 2022TransFuser: Imitation with Transformer-Based Sensor Fusion for Autonomous Drivingcited by 477
  10. 2020Learning Lane Graph Representations for Motion Forecastingcited by 380
  11. 2020Is Independent Learning All You Need in the StarCraft Multi-Agent Challenge?cited by 281
  12. 2020Emergent Complexity and Zero-shot Transfer via Unsupervised Environment Designcited by 260
13 more
  1. 2021TrafficSim: Learning to Simulate Realistic Multi-Agent Behaviorscited by 192
  2. 2022Evolving Curricula with Regret-Based Environment Designcited by 166
  3. 2021AdvSim: Generating Safety-Critical Scenarios for Self-Driving Vehiclescited by 132
  4. 2021Replay-Guided Adversarial Environment Designcited by 108
  5. 2022TrafficGen: Learning to Generate Diverse and Realistic Traffic Scenarioscited by 104
  6. 2022KING: Generating Safety-Critical Driving Scenarios for Robust Imitation via Kinematics Gradientscited by 95
  7. 2018GenAttack: Practical Black-box Attacks with Gradient-Free Optimizationcited by 82
  8. 2022GoRela: Go Relative for Viewpoint-Invariant Motion Forecastingcited by 68
  9. 2021Deep Multiagent Reinforcement Learning: Challenges and Directionscited by 42
  10. 2019Failure-Scenario Maker for Rule-Based Agent using Multi-agent Adversarial Reinforcement Learning and its Application to Autonomous Drivingcited by 36
  11. 2021Adversarial Reinforcement Learning for Procedural Content Generationcited by 36
  12. 2021Learning to Simulate Self-Driven Particles System with Coordinated Policy Optimizationcited by 33
  13. 2023Safety-Critical Scenario Generation Via Reinforcement Learning Based Editingcited by 16

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Automatic Curriculum Learning for Driving Scenarios: Towards Robust and Efficient Reinforcement Learning
  2. 2026Scaling Curriculum Learning For Autonomous Driving
  3. 2021Investigating Value of Curriculum Reinforcement Learning in Autonomous Driving Under Diverse Road and Weather Conditions
  4. 2025LearningFlow: Automated Policy Learning Workflow for Urban Driving with Large Language Models
  5. 2025CurricuVLM: Towards Safe Autonomous Driving via Personalized Safety-Critical Curriculum Learning with Vision-Language Models
  6. 2024Multi-Agent Reinforcement Learning for Autonomous Driving: A Survey
  7. 2025Scenario-Based Hierarchical Reinforcement Learning for Automated Driving Decision Making
  8. 2024Knowledge Transfer from Simple to Complex: A Safe and Efficient Reinforcement Learning Framework for Autonomous Driving Decision-Making