ScenarioNet

Open-source platform for large-scale traffic scenario modeling and simulation

ScenarioNet

Open-source platform for large-scale traffic scenario modeling and simulation

Meet ScenarioNet

ScenarioNet is an open-sourced platform for large-scale traffic scenario modeling and simulation.

  • ScenarioNet defines a unified scenario description format containing HD maps and detailed object annotations.
  • Scenarios recorded in this format can be replayed in the digital twins with multiple views, ranging from Bird-Eye-View layout to realistic 3D rendering.
  • ScenarioNet provides tools to build and manage databases built from various data sources including real-world datasets like Waymo, nuScenes, Lyft L5, and nuPlan datasets and synthetic datasets like the procedural generated ones and safety-critical ones.
  • We demonstrate several applications of ScenarioNet including large-scale scenario generation, AD testing, imitation learning, and reinforcement learning in both single-agent and multi-agent settings. The results imply scaling up the training data brings new research opportunities in machine learning and autonomous driving.
  • ScenarioNet's System Design

    Image
    ScenarioNet consists of the data layer, system layer, and application layer. Various datasets are unified into an internal scenario description. The system layer then provides a set of tools to operate on data efficiently, such as filtering, merging, sanity-check, splitting and so on. Once the database is ready, it can be loaded into MetaDrive for large-scale simulation and supports applications.

    Multi-sensory Simulation

    ScenarioNet leverages MetaDrive Simulator for multi-modal observation simulation.

    Traffic Scenarios from Various Datasets

    ScenarioNet reads from real world dataset such as Waymo, nuScenes, Lyft L5, and nuPlan datasets and creates interactive environment for closed-loop simulation.

    tSNE Visualization of Scenario Embeddings

    AD Stack Testing

    ScenarioNet bridges OpenPilot (Left) and ROS (Right) for autonomous driving testing.

    Acknowledgement

    This work was supported by the National Science Foundation under Grant No. 2235012 and the Samsung Global Collaboration Award.

    Reference

    @article{li2023scenarionet,
      title={ScenarioNet: Open-Source Platform for Large-Scale Traffic Scenario Simulation and Modeling},
      author={Li, Quanyi and Peng, Zhenghao and Feng, Lan and Liu, Zhizheng and Duan, Chenda and Mo, Wenjie and Zhou, Bolei},
      journal={Advances in Neural Information Processing Systems},
      year={2023}
    }