Aug 2026· International Conference on Electromechanical Control Technology and Transportation· Vol 14324, pp. 1432428 - 1432428-7· 0 citations· 11 references
Engineering
TL;DR
The results indicate that the multi-task CNN-LSTM can balance macroscopic behavior prediction and microscopic risk recognition, thereby improving the active warning capability of autonomous-driving systems in complex traffic scenarios.
Abstract
Accurate prediction of vehicle lane-change (LC) behavior and potential collision risk in highway scenarios is important for advanced driver-assistance systems (ADAS). To address the separation between lane-change prediction and risk assessment in existing studies, the difficulty of identifying long-tailed high-risk scenarios, and the tendency of selfattention models to overlook local high-frequency hazard signals, this paper proposes a multi-task learning model based on convolutional neural networks and long short-term memory networks (CNN-LSTM). The model simultaneously predicts vehicle lane-change intention, time to lane crossing (TTLC), and collision risk level. Based on HighD naturalistic driving data, kinematic and interaction features of the target vehicle and surrounding vehicles are extracted; 1D-CNN is used to extract local spatial features, and a unidirectional LSTM is combined to capture temporal dependencies. Experimental results show that the proposed model achieves a lane-change intention prediction accuracy of 94.49%; in the high-risk classification task, its recall reaches 65.52%, outperforming Transformer (44.83%) and SVM (24.14%). The results indicate that the multi-task CNN-LSTM can balance macroscopic behavior prediction and microscopic risk recognition, thereby improving the active warning capability of autonomous-driving systems in complex traffic scenarios.
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
Early recognition of lane-change intention is essential for proactive decision-making in autonomous driving and advanced driver assistance systems. This paper proposes a Dual Neural-Calibrated Interacting Multiple Model (DNC-IMM) that improves adaptability to driving context while preserving the probabilistic structure and interpretability of a conventional IMM. The proposed method encodes driving-context information, including target-vehicle motion, gaps to surrounding vehicles, and relative velocities, with a neural network that calibrates both the transition-probability matrix and measurement likelihoods. The final intention is determined from the calibrated IMM mode posterior rather than from a separate direct classifier. Experiments on the highD dataset demonstrate that the proposed method reliably recognizes lane-change intentions before lane crossing and provides particularly strong performance at the earlier 2-3 s prediction horizons.
OBJECTIVE
The right-turn risk of heavy vehicles at urban intersections has become a significant challenge for urban traffic safety management. However, most existing studies focus on static risk factors, lacking dynamic risk quantification and prediction for heavy vehicles in high-risk right-turn scenarios. This study aims to propose a dynamic risk quantification method and introduce an advanced deep learning model to predict driving risk effectively.
METHODS
An entropy weighting method based on sliding time windows was proposed to dynamically quantify the right-turn driving risk of heavy vehicles at intersections. An Informer model was introduced to predict driving risk, and observation, interval, and prediction time windows were defined to generate model inputs and outputs. To validate the model performance, the Long Short-Term Memory (LSTM), Convolutional Neural Network-Bidirectional Attention-LSTM (CNN-Bi-Attention-LSTM), and Transformer were employed for comparison.
RESULTS
The results showed that under a 1 s interval time window, the Informer model performed best with a 10 s observation window and a 5 s prediction window. It achieved a prediction precision of 82.69% and an inference time of only 0.014 ms per sample. Compared with other models, the Informer model significantly improved the accuracy and efficiency of long-term risk prediction.
CONCLUSIONS
This study provides an effective method for driving risk prediction for heavy vehicles at urban road intersections, addressing the limitations of traditional static risk analysis and providing technical support for active safety management.
Wenping Li, Bo Xu, Baikun Zhu et al.· Traffic Injury Prevention· 0 citations
Lane-change intention prediction is critical for intelligent vehicles, and driver decisions depend on the perception and processing of driving context information. Despite advances in deep learning in this domain, further exploration of the driving context processing remains essential. This study proposes a soft mask attention mechanism to adaptively enhance or suppress input features from target and surrounding vehicles. After that, the masked features are extracted using a heterogeneous information encoder, thereby differentiating between target and surrounding vehicle information processing. The encoded features are then integrated using a multi-head attention mechanism, and the lane change probabilities are output through convolution operations and decoder layers. Experiments demonstrate: (1) The proposed SMILE-LC (Soft Mask Information with Lane-based Encoding for Lane Change) achieves optimal prediction performance across all perception ranges with strong efficiency and generalization. (2) The soft mask attention mechanism intuitively reveals the information processing patterns in prediction model. The target vehicle's lateral acceleration is the most critical feature, while the position information of surrounding vehicles is more important than their velocity and acceleration. (3) The heterogeneous information encoder significantly improves lane change prediction performance, and the proposed lane-based encoding strategy outperforms other encoding architectures. These results can significantly advance the development of Advanced Driver Assistance Systems (ADAS) and enhance lane-change safety.
Heavy trucks experience persistent forward collision risks due to driver heterogeneity and limitations of fixed-threshold Forward Collision Warning (FCW) systems, which fail to adapt to dynamic driver behavior categories and multimodal contextual cues. This study proposes the Adaptive Time-to-Avoidance (A-TTC) framework, a driver-in-the-loop FCW framework developed using a naturalistic dataset of 3,519 video-verified FCW-triggered events from 569 heavy trucks. To capture behavioral variability, a Multimodal Temporal Alignment Neural Network (MMTANN) is introduced, explicitly synchronizing facial cues, road scenes, and vehicle dynamics to infer driver behavior categories (Distracted, Normal, Harsh) with 84.80% accuracy. Leveraging this behavior category awareness, a hybrid approach utilizes a Gradient Boosting Decision Tree (GBDT) for real-time reaction-time prediction and a convolutional-enhanced Transformer (C-Trans) for braking-distance estimation, reducing prediction error (RMSE) by 51.8% over standard LSTM. These parameters dynamically calibrate a personalized TTA threshold against the real-time predicted minimum Time-to-Collision (TTC). In a retrospective event-based evaluation on pre-triggered FCW logs, A-TTC achieved an overall accuracy of 81.42% and reduced the event-level nuisance-warning proportion to 13.74%, while maintaining a threat-event recall of 89.25%. This research provides a data-driven and driver-adaptive approach to enhancing the safety and personalization of commercial vehicle collision avoidance systems.
Yibing Wang, Qing He, Jingqiu Guo et al.· Accident Analysis and Preven...· 0 citations
Vehicle intention prediction is a pivotal aspect in the agility and safety of autonomous vehicles in all driving scenarios; if genuine enhancement of autonomous vehicles are required, we need to make them adopt human interpretation of driver's intention especially in cases that require a lot of human interaction as well as complex driving behaviors like the ones at intersections, roundabouts and emergency cases such as sudden stops where vehicle intention prediction helps in taking the correct evasive action within a real time period where every second of action makes an impact and can prevent a catastrophe from taking place. In the worst case, it helps minimize the damage and make safety a priority. Intention prediction can also be used to enhance trajectory prediction (intention conditioned trajectory prediction). In this study, The INTENT framework is proposed using LSTM model to predict the vehicle's intention at intersections 2 seconds ahead of the event occurrence to predict whether the cars in intersections are going straight, turning left, or turning right. Various model experiments and ablation study are thoroughly tested on InD dataset achieving 99.71% accuracy.