X-TRACK: Physics-Aware xLSTM for Realistic Vehicle Trajectory Prediction

2025-10-31

AI Summary

This work addresses two key limitations in vehicle trajectory prediction: insufficient modeling of long-term temporal dependencies and lack of physical plausibility. To this end, we introduce xLSTM—the first application of this architecture in traffic forecasting—and propose X-TRACK, a dynamics-aware sequence model that jointly integrates xLSTM’s exponential gating mechanism and extended memory modules with vehicle kinematic constraints. Extensive experiments on the highD and NGSIM datasets demonstrate that X-TRACK significantly outperforms state-of-the-art baselines—including LSTM, Transformer, and graph neural network approaches—in both predictive accuracy (ADE/FDE) and physical feasibility (e.g., realistic acceleration and curvature distributions). Notably, it exhibits superior stability and realism in long-horizon predictions beyond 5 seconds. This work establishes a new paradigm for physics-guided deep temporal modeling in autonomous driving and intelligent transportation systems.

Technology Category

• Computer Vision: Motion & Tracking
• Planning, Routing, and Scheduling: Model-Based Reasoning
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Recent advancements in Recurrent Neural Network (RNN) architectures, particularly the Extended Long Short Term Memory (xLSTM), have addressed the limitations of traditional Long Short Term Memory (LSTM) networks by introducing exponential gating and enhanced memory structures. These improvements make xLSTM suitable for time-series prediction tasks as they exhibit the ability to model long-term temporal dependencies better than LSTMs. Despite their potential, these xLSTM-based models remain largely unexplored in the context of vehicle trajectory prediction. Therefore, this paper introduces a novel xLSTM-based vehicle trajectory prediction framework, X-TRAJ, and its physics-aware variant, X-TRACK (eXtended LSTM for TRAjectory prediction Constraint by Kinematics), which explicitly integrates vehicle motion kinematics into the model learning process. By introducing physical constraints, the proposed model generates realistic and feasible trajectories. A comprehensive evaluation on the highD and NGSIM datasets demonstrates that X-TRACK outperforms state-of-the-art baselines.

Problem

• Predicting realistic vehicle trajectories using xLSTM
• Integrating vehicle motion kinematics into learning
• Generating feasible trajectories with physical constraints

Innovation

• Uses xLSTM for vehicle trajectory prediction
• Integrates vehicle motion kinematics constraints
• Generates realistic trajectories with physical feasibility