2026· Revista Internacional de Métodos Numéricos para Cálculo y Diseño en Ingeniería· 0 citations
TL;DR
An efficient Mamba-based feature extraction framework for jointly encoding vehicle trajectories and map information is proposed and achieves superior performance in terms of minADE, minFDE, and minMR, while maintaining high computational efficiency.
Abstract
Accurate prediction of future vehicle trajectories is essential for ensuring safety and reliable decision-making in autonomous driving systems. However, existing deep learning-based approaches exhibit several limitations. Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) struggle to effectively model long-term temporal dependencies and complex agent interactions, while Transformer-based architectures often suffer from high computational complexity and limited efficiency. To overcome these challenges, this paper proposes an efficient Mamba-based feature extraction framework for jointly encoding vehicle trajectories and map information. By leveraging state-space modeling and a selective scanning mechanism, the proposed approach effectively captures longrange dependencies and enhances the representation of complex traffic behaviors. Specifically, raw scene data are first normalized and embedded into a unified feature space. A Mamba Encoder is then employed to extract high-level features from historical vehicle trajectories and map elements. Subsequently, Vehicle-Vehicle and Vehicle-Map interaction modules are introduced to explicitly model dynamic interactions among traffic participants and between vehicles and the surrounding map. The resulting high-dimensional features are further fused using an additional Mamba Encoder, while a Global Interaction Module is designed to capture scenelevel dependencies. Finally, a Gated Recurrent Unit (GRU) decoder generates multi-modal future trajectory predictions. Experimental results on the Argoverse 1 dataset demonstrate that the proposed method achieves superior performance in terms of minADE, minFDE, and minMR, while maintaining high computational efficiency.OPEN ACCESS Received: 28/01/2026 Accepted: 16/04/2026
Comprehensive experiments on the NGSIM dataset validate the proposed model, demonstrating robust performance across structured highway driving scenarios and both the accuracy and computational efficiency of the proposed architecture.
Yang Li, Chengqian Jin, Zhikang Li· IEEE Access· 0 citations
Accurate vehicle trajectory prediction is essential for autonomous driving safety. However, existing data-driven models often ignore kinematic constraints, causing lateral jitter and trajectory distortion, while purely kinematics-based models lack flexibility in complex interactions. To address this, this paper presents a multimodal trajectory prediction method combining kinematics with dynamic interaction features. Operating in the Frenet coordinate system, the proposed model extracts historical features via a Bidirectional Gated Recurrent Unit (Bi-GRU) and utilizes an Adaptive Social Gating Network (ASGN) with multi-head attention to filter irrelevant interaction noise. This paper introduces 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. The model is trained using a composite loss function (Focal Loss and Best-of-K) to mitigate dataset long-tail distribution and trajectory divergence. Experiments on the HighD dataset show the proposed model achieves a minADE of 0.425 m and a minFDE of 0.955 m, outperforming baselines and reducing Lat-ADE by 53.9% compared to Social-GAN. These results confirm the model generates smoother, kinematically interpretable trajectories with higher accuracy in long-tail lane-changing scenarios.
Accurate, real-time motion forecasting is crucial for safe operation of autonomous vehicles. State-space models, such as Mamba, can model long-range dependencies with linear computational complexity, reducing sequence modeling overhead. However, Mamba-based motion forecasting methods still rely on Transformer-style decoding frameworks, which have high computational complexity and struggle to meet real-time requirements in complex traffic scenarios. To address this, this paper proposes a multimodal motion forecasting method IPDM, based on the intent-parametrized dynamics Mamba model. First, we design the decoupled intent-state parameterization module, which decomposes multimodal queries into intent and state queries, constructs specific state-space model parameters, and generates multimodal trajectories with linear computational complexity, thereby reducing computational complexity of the decoding stage. Furthermore, addressing the challenge that single-time-scale state-space models within the Mamba decoding framework struggle to balance long-term memory with instantaneous maneuvering responses, we propose a dual-scale controlled scan mechanism. This mechanism constructs short-term and long-term flows and employs intent-aware gated adaptive fusion to enhance forecasting performance while maintaining linear decoding properties. Experimental results on the Argoverse 1 and Argoverse 2 datasets demonstrate that IPDM achieves state-of-the-art performance in inference speed and parameter efficiency, outperforming most existing methods while reducing the number of parameters by over 40%. Note to Practitioners—In autonomous driving systems, vehicles must determine the likely trajectories of surrounding road users over the next few seconds within a very short timeframe to support path planning, obstacle avoidance, and safety decisions. However, in practical deployment, motion forecasting modules must not only ensure accuracy but also meet the strict constraints of onboard edge hardware regarding computational resources and inference latency. While many existing multimodal prediction methods offer good performance, their decoding processes typically involve high computational overhead, making it difficult to balance accuracy and real-time performance in complex traffic scenarios. This work is motivated by the need to address this engineering challenge. The method proposed in this paper separates the modeling of “how the vehicle might want to move” from “how the trajectory evolves over time.” By doing so, it reduces the computational burden of the decoding stage while maintaining multimodal prediction capabilities, thereby enhancing the method’s potential for deployment on resource-constrained platforms. Experimental results demonstrate that the method maintains strong prediction performance while reducing both model parameters and inference overhead. It should be noted that the method has primarily been validated on standard public datasets; for real-world deployment, further consideration must be given to the impacts of perception noise, map errors, and extreme traffic conditions. Beyond autonomous driving, this approach is expected to be applicable to delivery robots, roadside intelligent devices, and other embedded automation systems requiring rapid motion forecasting.
Jianhang Liu, Ang Liu, Mu Zhou et al.· IEEE Transactions on Automat...· 0 citations
Autonomous driving planning is a key component of IoT-enabled intelligent transportation systems, requiring vehicles to generate safe, efficient, and executable trajectories in complex urban environments from multi-source contextual information. While imitation learning (IL) has shown promise on large-scale datasets, IL-based planners still suffer from limited coverage of complex long-tail interactions, weak consistency with downstream constrained refinement, and insufficient use of high level scene semantics under real time constraints. To address these issues, this paper proposes a large language model (LLM) enhanced differentiable trajectory planning framework for IoT-enabled autonomous driving. Specifically, we introduce a surrounding agent centric data augmentation strategy to reorganize sur rounding agent trajectories as additional planning supervision, thereby improving the training distribution without collecting additional raw data. We further design a complexity-aware asyn chronous LLM-based semantic enhancement module to extract scene-related high-level semantic features with controlled online overhead. In addition, a differentiable optimization module is incorporated to refine generated trajectories with explicit residual penalties while backpropagating optimization gradients to the upstream planner. Experiments show that the proposed method achieves the best overall scores of 83.63 and 78.29 on the nuPlan closed-loop nonreactive and reactive Hard20 benchmarks, respectively, and CARLA-ROS tests further verify its online deployment and real time closed-loop execution capability.
Shihao Zhang, Jing Yang, Ziyu Song et al.· IEEE Internet of Things Jour...· 1 citation
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
Due to the absence of traffic signal control and the difficulty in accurately estimating the future movements of surrounding vehicles, autonomous vehicle decision-making faces challenges at unsignalized intersections. This study proposes a trajectory prediction-aided deep reinforcement learning framework. First, a composite prioritized replay mechanism is introduced into the Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithm, jointly considering temporal-difference error and reward-based event severity to enhance critical-experience reuse. Second, a convolutional multi-layer long short-term memory (CM-LSTM) model predicts surrounding-vehicle trajectories through convolutional local-motion encoding and stacked LSTM temporal modeling, and the predicted trajectories are incorporated into the deep reinforcement learning state representation. A multi-objective reward function is designed to balance collision avoidance, passing efficiency, lane keeping, and task completion. In CARLA go-straight and left-turn tests, CLS-TD3 achieves success rates of 93.8% and 90.2%, collision rates of 2.5% and 4.2%, and average passing times of 5.18 s and 5.58 s. Compared with TD3, the success rates increase by 6.3 and 8.6 percentage points, while average passing times decrease by 18.8% and 20.5%. These results demonstrate that the proposed framework improves the safety and crossing efficiency of autonomous vehicle decision-making at unsignalized intersections.
Shufeng Wang, Yuhang Wang, Yongxin Lei et al.· Machines· 0 citations