RITA: Boost Driving Simulators with Realistic Interactive Traffic Flow
Zhengbang Zhu,Shenyu Zhang,Yuzheng Zhuang,Yuecheng Liu,Minghuan Liu,Liyuan Mao,Ziqin Gong,Shixiong Kai,Qiang Gu,Bin Wang,Siyuan Cheng,Xinyu Wang,Jianye Hao,Yong Yu
DOI: https://doi.org/10.1145/3627676.3627681
2023-12-08
Abstract:High-quality traffic flow generation is the core module in building simulators for autonomous driving. However, the majority of available simulators are incapable of replicating traffic patterns that accurately reflect the various features of real-world data while also simulating human-like reactive responses to the tested autopilot driving strategies. Taking one step forward to addressing such a problem, we propose Realistic Interactive TrAffic flow (RITA) as an integrated component of existing driving simulators to provide high-quality traffic flow for the evaluation and optimization of the tested driving strategies. RITA is developed with consideration of three key features, i.e., fidelity, diversity, and controllability, and consists of two core modules called RITABackend and RITAKit. RITABackend is built to support vehicle-wise control and provide traffic generation models from real-world datasets, while RITAKit is developed with easy-to-use interfaces for controllable traffic generation via RITABackend. We demonstrate RITA's capacity to create diversified and high-fidelity traffic simulations in several highly interactive highway scenarios. The experimental findings demonstrate that our produced RITA traffic flows exhibit all three key features, hence enhancing the completeness of driving strategy evaluation. Moreover, we showcase the possibility for further improvement of baseline strategies through online fine-tuning with RITA traffic flows.
Artificial Intelligence,Multiagent Systems,Robotics