Electroencephalography (EEG) is widely utilized for train driver state detection due to its high accuracy and low latency. However, existing methods for driver status detection rarely use the rich physiological information in EEG to improve detection performance. Moreover, there is currently a lack of EEG datasets for abnormal states of train drivers. To address these gaps, we propose a novel transfer learning model based on a hybrid attention mechanism, named hybrid attention-based transfer learning network (HATNet). We first segment the EEG signals into patches and utilize the hybrid attention module to capture local and global temporal patterns. Then, a channel-wise attention module is introduced to establish spatial representations among EEG channels. Finally, during the training process, we employ a calibration-based transfer learning strategy, which allows for adaptation to the EEG data distribution of new subjects using minimal data. To validate the effectiveness of our proposed model, we conduct a multistimulus oddball experiment to establish a EEG dataset of abnormal states for train drivers. Experimental results on this dataset indicate that: 1) Compared to the state-of-the-art end-to-end models, HATNet achieves the highest classification accuracy in both subject-dependent and subject-independent tasks at 94.26% and 87.03%, respectively, and 2) The proposed hybrid attention module effectively captures the temporal semantic information of EEG data.

As a valuable tool for knowledge representation and reasoning, fuzzy Petri nets (FPNs) have obtained widespread application in many fields and achieved ideal results to some degree. However, current methods ignore the reliability of expert evaluations and the indirect effects among propositions, which may lead to decision-making errors or inaccurate evaluations. Thus, the indirect-effect-incorporated linguistic Z-number Petri nets (ILZPNs) are proposed in this paper. The linguistic Z-number is presented to capture knowledge information more comprehensively with its related fuzzy rules. The concepts of indirect effects and its aggression operators are designed to enhance the knowledge reasoning capability of ILZPN. Besides, the formal definition and the corresponding simulation algorithm of ILZPN are presented to conduct knowledge representation and reasoning. Subsequently, an empirical case, i.e., generalized eco-driving behavior evaluation, is applied to verify the proposed approach. In addition, comparison and sensitivity analysis are performed to monitor the robustness of the results. The results prove that this study offers a significant reference for the research into similar issues.

3D occupancy perception holds a pivotal role in recent vision-centric autonomous driving systems by converting surround-view images into integrated geometric and semantic representations within dense 3D grids. Nevertheless, current models still encounter two main challenges: modeling depth accurately in the 2D-3D view transformation stage, and overcoming the lack of generalizability issues due to sparse LiDAR supervision. To address these issues, this paper presents GEOcc, a Geometric-Enhanced Occupancy network tailored for vision-only surround-view perception. Our approach is three-fold: 1) Integration of explicit lift-based depth prediction and implicit projection-based transformers for depth modeling, enhancing the density and robustness of view transformation. 2) Utilization of mask-based encoder-decoder architecture for fine-grained semantic predictions; 3) Adoption of context-aware self-training loss functions in the pertaining stage to complement LiDAR supervision, involving the re-rendering of 2D depth maps from 3D occupancy features and leveraging image reconstruction loss to obtain denser depth supervision besides sparse LiDAR ground-truths. Our approach achieves State-of-the-Art performance on the Occ3D-nuScenes dataset with the least image resolution needed and the most weightless image backbone compared with current models, marking an improvement of 3.3% due to our proposed contributions. Comprehensive experimentation also demonstrates the consistent superiority of our method over baselines and alternative approaches. Our code is available athttps://github.com/world-executed/GEOcc.git

Prompt-tuning has demonstrated impressive performance in continual learning by querying relevant prompts for each input instance, which can avoid the introduction of task identifier. Its forgetting is therefore reduced as this instance-wise query mechanism enables us to select and update only relevant prompts. In this paper, we further integrate prompt-tuning with gradient projection approach. Our observation is: prompt-tuning releases the necessity of task identifier for gradient projection method; and gradient projection provides theoretical guarantees against forgetting for prompt-tuning. This inspires a new prompt gradient projection approach (PGP) for continual learning. In PGP, we deduce that reaching the orthogonal condition for prompt gradient can effectively prevent forgetting via the self-attention mechanism in vision-transformer. The condition equations are then realized by conducting Singular Value Decomposition (SVD) on an element-wise sum space between input space and prompt space. We validate our method on diverse datasets and experiments demonstrate the efficiency of reducing forgetting both in class incremental, online class incremental, and task incremental settings. The code is available at https://github.com/JingyangQiao/prompt-gradient-projection.

Combining LiDAR points and images for robust semantic segmentation has shown great potential. However, the heterogeneity between the two modalities (e.g. the density, the field of view) poses challenges in establishing a bijective mapping between each point and pixel. This modality alignment problem introduces new challenges in network design and data processing for cross-modal methods. Specifically, 1) points that are projected outside the image planes; 2) the complexity of maintaining geometric consistency limits the deployment of many data augmentation techniques. To address these challenges, we propose a cross-modal knowledge imputation and transition approach. First, we introduce a bidirectional feature fusion strategy that imputes missing image features and performs cross-modal fusion simultaneously. This allows us to generate reliable predictions even when images are missing. Second, we propose a Uni-to-Multi modal Knowledge Distillation (U2MKD) framework, leveraging the transfer of informative features from a single-modality teacher to a cross-modality student. This overcomes the issues of augmentation misalignment and enables us to train the student effectively. Extensive experiments on the nuScenes, Waymo, and SemanticKITTI datasets demonstrate the effectiveness of our approach. Notably, our method achieves an 8.3 mIoU gain over the LiDAR-only baseline on the nuScenes validation set and achieves state-of-the-art performance on the three datasets.

Detecting driver fatigue is critical to ensuring road safety. Existing fatigue detection methods typically rely on traditional hand-picked features as inputs. However, these hand-picked features can hardly respond accurately to the driver’s fatigue state due to a certain degree of subjectivity and the extraction of these features requires a long time window, which limits the accuracy and real-time performance of the detection. This paper proposes a novel fatigue detection method based on multi-source information fusion, which relies entirely on neural networks for automatic feature extraction. Through simulated driving experiments, we recorded physiological signals and facial videos from 21 participants for model training and testing. The results show that our model outperforms existing methods in terms of accuracy and real-time performance, achieving a detection accuracy of 93.15% within a 3-second time window (specificity = 94.04%, sensitivity = 91.71%). The visualization results of the model reveal potential relationships between facial regions for the first time, validating the rationality and effectiveness of our method. The practical issues of fatigue detection methods and future research directions are also explored.

The human–rail vehicle transportation system safety design is complicated given that the complexity of the multilevel system with parameter uncertainties propagating from the vehicle structure (primary collision) to the interior human compartment (secondary collision). This article establishes a hybrid framework incorporating a stochastic approach and an integrated optimization strategy to improve train crashworthiness and reduce passenger crash injuries. The stochastic approach utilizes adaptive sparse polynomial chaos expansion models and variance-based sensitivity indices to evaluate the statistic characteristics of system responses and quantify the contribution ranking of uncertain parameters to response variations. The optimization strategy integrating the evolutionary algorithm and the multicriteria decision making is proposed to solve the nonuniqueness of Pareto optimal solutions. In the optimization process, the modified DEMATEL–ANP method with interval type-2 fuzzy sets is developed to deal with vague linguistic judgments for the importance sequence of human injury responses. Theq-rung orthopair trapezoidal fuzzy uncertain linguistic sets–TOPSIS method is established to address hesitant linguistic evaluations for the Pareto front and select the final optimal solution. Compared with the initial design, the driver abbreviated injury scale (AIS) 3 plus joint injury probability is reduced from 67.08% to 14.17% after optimization. Results prove that the proposed framework is a practical tool for improving the passive safety of railway industry.

Train driver fatigue and distraction are the main reasons for railway accidents. One of the new technologies to monitor drivers is by using the EEG signals, which provides vital signs monitoring of fatigue and distraction. However, monitoring systems involving full-head scalp EEG are time-consuming and uncomfortable for the driver. The aim of this study was to evaluate the suitability of recently introduced forehead EEG for train driver fatigue and distraction detection. We first constructed a unique dataset with experienced train drivers driving in a simulated train driving environment. The EEG signals were collected from an EEG recording device placed on the driver’s forehead, and numerous features including energy, entropy, rhythmic energy ratio and frontal asymmetry ratio were extracted from the EEG signals. Therefore, a time-series ensemble learning method was proposed to perform fatigue and distraction detection based on the extracted feature. The proposed method outperforms other popular machine learning algorithms including Support Vector Machine(SVM), K-Nearest Neighbor(KNN), Decision Tree(DT), and Long short-term memory(LSTM). The proposed method is stable and convenient to meet the real-time requirement of train driver monitoring.

With the increasing speed of railway vehicles, deciding how to reasonably distribute impact energy to each vehicle has been a widespread concern in safety protection systems. This article formulates a three-dimensional train-track coupling dynamics model using MAthematical DYnamic MOdels (MADYMO) multibody dynamics software. A train-to-train collision is then simulated using this model. A hybrid solution methodology that combines the non-dominated sorting genetic algorithm II (NSGA-II), modified best and worst method with cloud model theory and grey relational analysis is proposed. The optimization parameters and objectives are determined based on the EN15227 crashworthiness requirements for railway vehicles. An empirical case of an existing train with eight vehicles that have been in operation in China is applied to verify this dynamics model derived from a high-speed train and solution methodology. Analysis and discussion are conducted to monitor the robustness of the results and the practical implications for rail transportation are summarized. The results prove that the obtained optimal solution by this research has better crashworthiness than an existing solution.