Jul 2026· Signal Processing and Communications Applications Conference· pp. 1-4· 0 citations· 21 references
Abstract
Safe decision making in autonomous driving relies on accurately predicting the future positions of traffic actors. Although existing trajectory prediction methods mainly focus on map information and social interactions, actors’ intrinsic motion dynamics are often overlooked. In this study, we propose a dynamic trajectory prediction architecture based on velocity profiles extracted from observed past trajectories. The proposed architecture decomposes driving scenes into low- and high-velocity models and routes each sample to the corresponding expert model. Experiments on a large-scale real-world autonomous driving dataset show that velocity-based decomposition reduces prediction error compared with conventional single-model baselines.
A Stochastic Gating Decoder for multimodal latent variable sampling, adaptively fusing kinematics and data-driven paths to capture driver intention uncertainty while maintaining kinematic consistency is introduced.
A comprehensive Explainable AI (XAI) evaluation framework is introduced, including temporal sensitivity analysis, interaction-aware perturbation studies, spatial influence analysis, and gradient-based feature attribution methods that provide insights into how the model captures temporal motion dependencies, neighboring vehicle interactions, and environmental context during trajectory prediction.
R. Megalingam, Naveen Prasaad Selvarajan, Pritty Vijay· Italian National Conference...· 0 citations
Accurate short-horizon vehicle trajectory prediction is critical for autonomous driving and vehicle-to-everything (V2X) communication. Most existing deep learning-based prediction methods rely on fixed-length historical trajectories and multi-agent context, which may be unavailable when a vehicle is newly observed. This paper studies single-frame short-horizon trajectory prediction at signalized intersections, where only one enriched observation frame is used for future motion-state forecasting. We propose a Neural Ordinary Differential Equation (Neural ODE)-based framework that formulates single-frame prediction as continuous-time motion-state forecasting. The proposed model encodes instantaneous kinematic variables, road-context information, and training-set-derived spatial priors into a latent representation, evolves the latent state in continuous time, and decodes future state variables. The predicted motion states are converted into future vehicle coordinates through kinematic integration. Experiments on the CitySIM-Intersection A dataset, evaluated by ADE and DE, show that the proposed method achieves competitive short-horizon prediction accuracy under the single-frame setting. The experiments include baseline comparisons with classical motion models and single-frame neural baselines, input ablation, coordinate and spatial-cell analyses, runtime evaluation, and supplementary studies on multi-frame sequence baselines, longer-horizon rollout, robustness, maneuver-specific performance, and statistical significance. The results clarify the applicability and limitations of single-frame short-horizon prediction on both straight and curved driving subsets.
Yijun Tang, Wenhao Huang, Yang Pu et al.· Discover Computing· 0 citations
Accurate trajectory prediction is critical for autonomous driving. To address extended-horizon error accumulation during nonlinear highway lane changes, the HGCA-Net framework is proposed. It employs a dynamic spatio-temporal graph attention network for interaction modeling and an intention recognition branch for maneuver preferences. A cross-attention mechanism integrates these priors to alleviate intention ambiguity, while a time-weighted loss penalizes far-end drift. Experiments on the NGSIM dataset demonstrate that HGCA-Net reduces ADE and FDE to 1.9632 m and 4.7359 m, outperforming baselines at the 5-second horizon.
Z. Wang, Ge Tong, Tao Cheng et al.· International Conference on...· 0 citations
Trajectory planning for connected and automated vehicles requires both scene-aware decision making and multimodal future trajectory generation. However, directly generating trajectories from a single standard noise distribution may lead to insufficient mode separation and reduced candidate diversity. To address this issue, this paper proposes a MeanFlowbased trajectory planning framework that introduces trajectory-cluster-guided Gaussian noise modeling into the generation process. Human driving trajectories are first clustered into multiple behavior modes, and each mode is represented by an individual Gaussian noise distribution. A scene encoder then extracts contextual information from ego-vehicle history, neighboring vehicles, and lane-level features. Conditioned on the scene context and trajectory mode, the MeanFlow decoder generates future trajectory candidates, from which the final trajectory is selected according to the predicted mode probability. Experiments under different mode settings demonstrate that the proposed framework can generate feasible multimodal trajectory candidates and select stable planning outputs for CAV trajectory planning.
Dongting Ge· International Conference on...· 0 citations
Deep learning-based vessel trajectory prediction using Automatic Identification System (AIS) has become a hot topic in the fields of maritime traffic monitoring, situational awareness, and navigational decision support. However, most previous studies have focused primarily on end-to-end model training using trajectory data from a single water area, which limits the resulting models’ ability to generalize to regions with different traffic patterns. To address this issue, this study proposes a method that constructs traffic context from historical AIS records at multiple geographic resolutions using H3, a hexagonal hierarchical spatial indexing system, and integrates this context with a Transformer-based trajectory predictor. A reliability-aware selector determines the contribution of the context to the final prediction, conditioning this decision on the vessel’s motion state and the retrieved historical patterns. Experiments on AIS data from three distinct water areas demonstrated that H3-indexed context improved cross-water prediction accuracy without requiring model retraining on the target area. These findings demonstrate that H3-indexed context, structured at multiple geographic resolutions and integrated through a selective mechanism, serves as transferable spatial context for vessel trajectory prediction.
