Multi-Agent Driving Behavior Prediction across Different Scenarios with Self-Supervised Domain Knowledge

How to make precise multi-agent trajectory prediction is a crucial problem in the context of autonomous driving. It is significant to have the ability to predict surrounding road participants' behaviors in many different, seen or unseen scenarios for enhancing autonomous driving safety and efficiency. Extensive research has been conducted to improve the overall prediction performance based on one enormous dataset or pay attention to some specified scenarios. However, how to generalize the prediction to different scenarios is less investigated. In this paper, we introduce a graph-neural-network-based framework for multi-agent interaction-aware trajectory prediction. In contrast to recent works which use the Cartesian coordinate system and global context images directly as input, we propose to leverage human's prior knowledge such as the comprehension of pairwise relations between agents and pairwise context information extracted by self-supervised learning approaches to attain an effective Frenét-based representation. We evaluate our method across different traffic scenarios with diverse layouts and compare it with state-of-the-art methods. We demonstrate that our approach achieves superior performance in terms of overall performance, zero-shot and few-shot transferability.