🤖 AI Summary
To address the low efficiency and poor scalability of manually constructing diverse, complex multi-agent interaction scenarios for autonomous driving, this paper proposes MATSGym—a CARLA-based multi-agent traffic scenario generation framework. Methodologically, MATSGym unifies heterogeneous traffic scenario specifications for the first time, integrating partially observable scenario modeling with unsupervised environment design (UED) to automatically synthesize scalable, heterogeneous multi-agent interaction scenarios from sparse, high-level specifications. It further incorporates a dynamic adaptive curriculum learning mechanism to accelerate training and improve policy robustness. Experimental results demonstrate that policies trained on MATSGym-generated scenarios achieve significantly enhanced generalization and safety in complex interactive driving environments. The framework is open-sourced, enabling flexible extension and community reproducibility.
📝 Abstract
The automated generation of diverse and complex training scenarios has been an important ingredient in many complex learning tasks. Especially in real-world application domains, such as autonomous driving, auto-curriculum generation is considered vital for obtaining robust and general policies. However, crafting traffic scenarios with multiple, heterogeneous agents is typically considered as a tedious and time-consuming task, especially in more complex simulation environments. In our work, we introduce MATS-Gym, a Multi-Agent Traffic Scenario framework to train agents in CARLA, a high-fidelity driving simulator. MATS-Gym is a multi-agent training framework for autonomous driving that uses partial scenario specifications to generate traffic scenarios with variable numbers of agents. This paper unifies various existing approaches to traffic scenario description into a single training framework and demonstrates how it can be integrated with techniques from unsupervised environment design to automate the generation of adaptive auto-curricula. The code is available at https://github.com/AutonomousDrivingExaminer/mats-gym.