Towards Human-Like Trajectory Prediction for Autonomous Driving: A Behavior-Centric Approach

2025-05-27

AI Summary

To address low prediction accuracy and poor behavioral interpretability in vehicle trajectory forecasting under complex dynamic traffic scenarios, this paper proposes HiT, a behavior-centric model. HiT innovatively introduces dynamic graph centrality measures to model both direct and higher-order indirect interactions among vehicles. It further incorporates a behavior-aware embedding module and a dynamic graph neural network, overcoming the limitations of static graph modeling to enable robust representation and interpretable prediction of fine-grained driving behaviors (e.g., aggressive maneuvering). Evaluated on five large-scale real-world datasets—NGSIM, HighD, RounD, ApolloScape, and MoCAD++—HiT consistently outperforms state-of-the-art methods. Notably, under aggressive driving scenarios, it achieves average displacement error (ADE) and final displacement error (FDE) reductions of 12.3% and 14.8%, respectively.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Planning, Routing, and Scheduling: Model-Based Reasoning
• Computer Vision: Motion & Tracking

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Predicting the trajectories of vehicles is crucial for the development of autonomous driving (AD) systems, particularly in complex and dynamic traffic environments. In this study, we introduce HiT (Human-like Trajectory Prediction), a novel model designed to enhance trajectory prediction by incorporating behavior-aware modules and dynamic centrality measures. Unlike traditional methods that primarily rely on static graph structures, HiT leverages a dynamic framework that accounts for both direct and indirect interactions among traffic participants. This allows the model to capture the subtle yet significant influences of surrounding vehicles, enabling more accurate and human-like predictions. To evaluate HiT's performance, we conducted extensive experiments using diverse and challenging real-world datasets, including NGSIM, HighD, RounD, ApolloScape, and MoCAD++. The results demonstrate that HiT consistently outperforms other top models across multiple metrics, particularly excelling in scenarios involving aggressive driving behaviors. This research presents a significant step forward in trajectory prediction, offering a more reliable and interpretable approach for enhancing the safety and efficiency of fully autonomous driving systems.

Problem

• Enhancing trajectory prediction for autonomous driving using behavior-aware modules
• Capturing subtle vehicle interactions via dynamic centrality measures
• Improving prediction accuracy in aggressive driving scenarios

Innovation

• Behavior-aware modules enhance trajectory prediction
• Dynamic centrality measures capture vehicle interactions
• Outperforms top models in aggressive driving scenarios