Reading power equipment meters often requires loads of manpower, which is a trivial, repetitive, and error-prone task. While conventional automated recognition methods using computer vision (CV) techniques are inflexible under diverse scenarios, in this article, we propose a lightweight meter recognition method that combines deep learning and traditional CV techniques for automated meter reading. For meter detection, an adaptive anchor and global context (GC) module are deployed to improve the feature extraction ability of lightweight backbone without increasing computational cost. Then, an feature pyramid network (FPN) and a path aggregation network (PANet) are developed to realize the information interaction between different feature layers and achieve multiscale prediction. Our method also includes a multitask segmented network to read the detected meters, accelerating the detection speed. Experiments demonstrate that our proposed method can achieve a detection speed of 123 frame per second (FPS) in GeForce GTX 1080 and can obtain an accuracy of 88.2% mean average precision (mAP)50:95. In the case of insufficient training samples, the method can still achieve an accuracy of 80.9% mAP50:95. In addition, we build a power meter images (PMIs) dataset, which contains 1800 images in real scene. The dataset and method we proposed can help with further upgrades of traditional substations. In the future, we also hope to extend the algorithm to edge computing cameras for substations. The newly developed dataset and code are available at https://github.com/zzfan3/electric_meter_detect_recognize .

Pedestrian intrusion detection is crucial to ensuring safe railway operation. Current pedestrian detection algorithms lack consideration for real-world railway scenarios, such as the reflective properties of screen doors and train windows, may mistakenly trigger pedestrian intrusion alerts. Scale variability and pedestrian overlap often lead to detection inaccuracy, making them inadequate for addressing the specific requirements of railway perimeter security. This letter introduces an innovative pedestrian detection algorithm that incorporates Mirror Translation Attention (MTA) and Feature Pooling Enhancement (FPE). MTA, including mirror flipping and offsetting the feature mapping, could significantly mitigate missed detection caused by reflective surfaces. Additionally, we introduce sparsity to the inputs of the self-attention, which significantly enhancing the model's inference speed. A multi-scale approach is adopted to accommodate the diversity in pedestrian sizes, while the FPE addresses occlusion issues across various scales. Compared to the advanced YOLOv8 model, the proposed method improves AP50 by 1.6% to 92.11% and reduces model parameters by 63.55% in our self-built railway pedestrian intrusion dataset.

To overcome the limitations of current methods in architectural scene segmentation accuracy, this study presents a hybrid attention-enhanced YOLOv8 framework and introduces a dedicated building interior profile dataset. The proposed approach extends the YOLOv8 architecture by integrating SimAM (Simple, Parameter-Free Attention Module) to dynamically evaluate neuron significance and refine feature representations. This is coupled with an Efficient Multi-Scale Attention (EMA) module that synergizes local and global attention mechanisms, enabling robust multi-scale feature fusion while maintaining stable weight optimization. To address the scarcity of domain-specific data, a meticulously annotated dataset encompassing common architectural interior elements is developed. Comprehensive evaluations demonstrate that the enhanced model achieves an 89.9% mAP@0.5 on the proposed dataset, outperforming the baseline YOLOv8 with relative improvements of 4.5% in precision, 5.2% in recall, 5.1% in mAP@0.5, and 4.5% in mAP@0.5–0.95. These advancements underscore the efficacy of hybrid attention mechanisms in architectural scene analysis and establish a benchmark dataset to facilitate future research in intelligent building environment interpretation.

With the rapid development of new energy vehicle technologies, higher demands have been placed on fault diagnosis for automotive transmission gearboxes. To address the poor adaptability of traditional methods under complex operating conditions, this paper proposes a sensor data-driven fault diagnosis method based on improved ensemble empirical mode decomposition (EEMD) combined with convolutional neural networks (CNNs) and Bidirectional Long Short-Term Memory (BiLSTM) networks. The method incorporates a dynamic noise adjustment mechanism, allowing the noise amplitude to adapt to the characteristics of the signal. This improves the stability and accuracy of signal decomposition, effectively reducing the instability and error accumulation associated with fixed-amplitude white noise in traditional EEMD. By combining the CNN and BiLSTM modules, the approach achieves efficient feature extraction and dynamic modeling. First, vibration signals of the transmission gearbox under different operating states are collected via sensors, and an improved EEMD method is employed to decompose the signals, removing background noise and nonstationary components to extract diagnostically significant intrinsic mode functions (IMFs). Then, the CNN is utilized to extract features from the IMFs, deeply mining their spatiotemporal characteristics, while the BiLSTM captures the temporal sequence dependencies of the signals, enhancing the comprehensive modeling of nonlinear and dynamic fault features. The combination of these two networks enables efficient adaptation to complex conditions, achieving accurate classification and identification of multiple gearbox fault modes. Results indicate that the proposed approach is highly accurate and robust for identifying gearbox fault modes, significantly exceeding the performance of conventional methods and isolated network models. This provides an efficient and intelligent solution for fault diagnosis of automotive transmission gearboxes.