SIPTraj: Map-Free End-to-End Trajectory Prediction via Physics-Guided Scene Interaction

2026-08-01

AI Summary

This work addresses the challenge of predicting physically plausible trajectories for traffic agents without reliance on high-definition maps, a setting where existing methods struggle to effectively model interactions between agents and scene context while ensuring kinematic feasibility. To this end, we propose SIPTraj, a novel framework comprising a Hierarchical Agent-Scene Encoder (HASE) and a Physics-Guided Iterative Decoder (PGID). HASE anchors agents to scene evidence through an agent-guided interaction mechanism, while PGID enforces physical plausibility by incorporating kinematic state supervision directly into the internal representation during decoding. Extensive experiments demonstrate that SIPTraj significantly outperforms both map-free approaches and strong map-dependent baselines on the nuScenes and Argoverse 2 Sensor datasets, achieving superior prediction accuracy and enhanced physical consistency.

Abstract

Trajectory prediction of surrounding agents is a prerequisite for safe planning and decision making in autonomous driving. Without high-definition (HD) maps, sensor-derived bird's-eye-view (BEV) features provide no explicit lane topology or drivable-area priors, making it inherently difficult to ground each agent in its surrounding scene context. Moreover, physical feasibility remains difficult to capture through data-driven learning alone, as kinematic constraints on agent motion cannot be explicitly encoded without structured supervision. Existing map-free predictors extract scene context in an agent-agnostic manner through a single fusion step and treat physical constraints only as output-level penalties, leaving both challenges unaddressed. We propose SIPTraj, a map-free trajectory prediction framework that jointly addresses scene grounding and physical feasibility. SIPTraj introduces a Hierarchical Agent-Scene Encoder (HASE) progressively grounding each agent in agent-guided scene evidence and refining inter-agent relations within the scene-grounded space. To tackle physical infeasibility in predicted trajectories, we develop a Physics-Guided Iterative Decoder (PGID). It conditions decoding on instantaneous kinematic states, propagating physical supervision into internal representations rather than output trajectories alone. Extensive experiments on nuScenes and Argoverse 2 Sensor show that SIPTraj surpasses prior map-free predictors and strong map-based baselines without any HD map at inference. Our code will be released as open-source.

Problem

• trajectory prediction
• map-free
• scene interaction
• physical feasibility
• autonomous driving

Innovation

• map-free trajectory prediction
• scene grounding
• physics-guided decoding
• agent-scene interaction
• kinematic constraints