High-speed railways are widely used worldwide due to their high speeds, punctuality, and comfort. Autonomous operation and automatic obstacle avoidance technologies are the key to enhancing operational efficiency and ensuring passenger safety. By identifying the tracks, high-speed rail clearance can be effectively detected, which helps to determine the specific space through which the train travels. In this paper, a lightweight track line tracking model (Track${}^{\mathbf {2}}$Net) is proposed, which is based on keypoint alignment and track anchor identification. Firstly, the method mathematically analyses the shapes of tracks in non-turnout scenes and then implements a rapid track matching algorithm based on keypoint alignment. For complex track shapes that fail to match, the Complex Track Positioning Module (CTPM) is proposed. The module extracts features by pre-defining the track anchor lines and has better robustness in complex scenes. Additionally, a post-treatment track method is proposed for turnouts to further refine the detection of the track segment occupied by the train. In order to judge the accuracy of track detection more rigorously, improved evaluation metrics are proposed. On the track datasets of the train operation constructed in this paper, Track${}^{\mathbf {2}}$Net achieves an accuracy of 91.43% and processes images at a rate of 236 FPS. Experiments verify the effectiveness of the model in track line detection.

Recent advances in generative models have made it increasingly difficult to distinguish real data from model-generated synthetic data. Using synthetic data for successive training of future model generations creates “self-consuming loops,” which may lead to model collapse or training instability. Furthermore, synthetic data is often subject to human feedback and curated by users based on their preferences. Ferbach et al. (2024) recently showed that when data is curated according to user preferences, the self-consuming retraining loop drives the model to converge toward a distribution that optimizes those preferences. However, in practice, data curation is often noisy or adversarially manipulated. For example, competing platforms may recruit malicious users to adversarially curate data and disrupt rival models. In this paper, we study how generative models evolve under self-consuming retraining loops with noisy and adversarially curated data. We theoretically analyze the impact of such noisy data curation on generative models and identify conditions for the robustness and stability of the retraining process. Building on this analysis, we design attack algorithms for competitive adversarial scenarios, where a platform with a limited budget employs malicious users to misalign a rival’s model from actual user preferences. Experiments on both synthetic and real-world datasets demonstrate the effectiveness of the proposed algorithms.

The pantograph slider is a key component of the pantograph-catenary system. It is important to monitor the wear of sliders for rail transit safety. In this paper, an innovative real-time high-precision lightweight approach is proposed to estimate the wear of the slider. It allows complete monitoring of all sliders of the pantograph. In the first stage, a method based on image processing and object detection by deep learning is proposed to locate the region of the slider. It takes into account the large aspect ratio on the pantograph slider and the inclined angle. In the second stage, the neural network for wear estimation of pantograph slider (WEPSNet) is proposed. It realizes end-to-end contour extraction of the slider. The residual thickness of the slider is calculated by counting the number of pixels and the error is analyzed. Furthermore, the error arising from the perspective projection transformation in the monocular image is discussed. The experimental results demonstrate that, with the similar model size, the proposed WEPSNet outperforms the state-of-the-art method by 1.08% mIoU and 4.63% IMP. Moreover, the accuracy of residual thickness is tested on 120 pantograph slider images, achieving up to 95.91% within the allowable 1mm error, which is 6.68% higher than the state-of-the-art method.

Complex multistate systems (MSSs) are composed of components with multiple levels of performance and complex internal component structure, and they are widely used in engineering modeling. Failure of complex MSSs can cause significant economic losses and even safety issues. Therefore, optimizing the maintenance strategy to ensure reliable operation is a crucial issue. This article investigates the problem of maintenance strategy optimization for complex MSSs. First, the phase-type distribution is introduced to construct the infinitesimal generator matrix of components with multiple level maintenance considered. Second, the universal generation function method is employed to build the performance function of the system. After that, the i-NSGA-Ⅱ-dc algorithm is proposed to solve the maintenance strategy optimization problem with the biobjective of minimizing expense ratio and maximizing expected performance. Finally, the fuzzy-logic-based approach is used to assist decision-makers in selecting preferred solutions. The door system of the intercity multiple units is taken as an example to illustrate the applicability of the approach, and the results show that the door system is excessively maintained currently. The simulation results indicate that the optimized maintenance strategy can extend the maintenance interval and reduce the frequency of maintenance, and lower the maintenance cost rate in the meantime.

To accomplish computationally intensive visual tasks on mobile devices with limited memory and computation capability, a common solution is to offload the tasks to the edge with more powerful computing resources. Nevertheless, due to the extra communication, even edge computing is challenged by the growing demands for real-time interactive tasks (e.g., VR and AR), which are generally performed by deep CNNs with heavy computation. In this paper, we design a novel system called HRCache to reduce the end-to-end latency by effectively recompressing video data and reusing cached inference results on the edge. As similar image regions have been indicated in the offloading video coding, the edge can approximately infer these regions with results stored on the edge to save the time for calculating these regions. Moreover, since the data in these regions do not require calculation, the transmitted data can be further simplified to reduce transmission latency. Furthermore, HRCache can quickly and continuously adjust the coding parameters to adapt the accuracy loss and overall latency with application requirements. Compared with the original offloading schemes, HRCache significantly reduces the average latency, about 13.60% to 18.83%, at little accuracy loss of 1.25% in Top-1 accuracy for classification and 0.135 in IoU for object detection.

Pantograph is one of the most important components in electrical railway vehicles. To guarantee steady power supply for the train, the surface of the pantograph slide plate should be smooth enough so that the catenary can move on it from one side to the other side steadily with low friction. In addition, the thickness of the pantograph slide plate cannot be smaller than the lower limit for the sake of safety. Therefore, periodical inspection and maintenance of the pantograph slide plate are significant in terms of safe and stable operation. In this paper, an innovative and intelligent method based on deep learning and image processing technologies is proposed for the online condition monitoring of the pantograph slide plate. In the first stage, the surface defect detection and recognition method of the pantograph slide plate is proposed. Four typical surface defects of the slide are considered, and a deep learning model, pantograph defect detection neural network (PDDNet), is trained for the defect detection and recognition. In the second stage, five key criteria for qualifying the wear condition are proposed. The wear edge estimation based on image processing technology is investigated in detail. Furthermore, they are used to calculate the wear depth and evaluate the wear condition of the pantograph slide. The experiment results demonstrate that the proposed PDDNet can detect the surface defects and also recognize the four kinds of defects with a sound accuracy. The wear depth estimation results are compared with on-site measurement data, and the proposed method can achieve high estimation accuracy.