KI-GAN: Knowledge-Informed Generative Adversarial Networks for Enhanced Multi-Vehicle Trajectory Forecasting at Signalized Intersections

TL;DR

This work tackles the challenge of predicting vehicle trajectories at signalized intersections, where complex road geometry, traffic signals, and multi-vehicle interactions complicate forecasting. The authors introduce KI-GAN, a knowledge-informed GAN framework that fuses four encoders (Trajectory, Motion, Physical Attributes, Traffic) with a specialized Vehicle Attention Pooling Net (VAP-Net) and an LSTM-based decoder, guided by a GAN discriminator. Key contributions include the multi-encoder architecture, the VAP-Net pooling mechanism, and robust evaluation on the SinD dataset, showing state-of-the-art ADE/FDE performance for 6-second and 9-second prediction horizons. The results demonstrate the practical impact of incorporating traffic-signal information and nuanced vehicle interactions for accurate trajectory forecasting at intersections, with potential benefits for traffic management and autonomous driving systems.

Abstract

Reliable prediction of vehicle trajectories at signalized intersections is crucial to urban traffic management and autonomous driving systems. However, it presents unique challenges, due to the complex roadway layout at intersections, involvement of traffic signal controls, and interactions among different types of road users. To address these issues, we present in this paper a novel model called Knowledge-Informed Generative Adversarial Network (KI-GAN), which integrates both traffic signal information and multi-vehicle interactions to predict vehicle trajectories accurately. Additionally, we propose a specialized attention pooling method that accounts for vehicle orientation and proximity at intersections. Based on the SinD dataset, our KI-GAN model is able to achieve an Average Displacement Error (ADE) of 0.05 and a Final Displacement Error (FDE) of 0.12 for a 6-second observation and 6-second prediction cycle. When the prediction window is extended to 9 seconds, the ADE and FDE values are further reduced to 0.11 and 0.26, respectively. These results demonstrate the effectiveness of the proposed KI-GAN model in vehicle trajectory prediction under complex scenarios at signalized intersections, which represents a significant advancement in the target field.

Figures (1)

• Figure 1: Architecture of Knowledge-Informed Generative Adversarial Networks