Synchronized Trajectory Prediction for Hybrid Multi-agent via Attention-Denoised Endpoint Distribution

Trajectory prediction is critical for applications related to connected autonomous vehicles (CAVs), where multi-agent trajectory prediction can significantly reduce collisions and congestion in highways or hybrid-open scenarios. It serves as the foundation for autonomous driving, enabling vehicles to navigate complex environments safely and efficiently. Prior methods have assessed the spatio-temporal dynamics of agents but often neglected intrinsic intent and uncertainty, thereby limiting their effectiveness. We present the Denoised Endpoint Distribution model for trajectory prediction, which distinctively models agents’ spatio-temporal features alongside their intrinsic intentions and uncertainties. By employing Diffusion and Transformer models to focus on agent endpoints rather than entire trajectories, our approach significantly reduces model complexity and enhances performance through endpoint information. In addition, the designed attention-aware spatio-temporal graphs provide strong guidance information for the diffusion model, enhancing its ability to accurately predict trajectories. Our experiments on open datasets, including highways and hybrid open scenarios, along with comparative and ablation studies, demonstrate the validity of our model and the importance of its components.