Search Results (240)

Automated surface defect detection in steel manufacturing is pivotal for ensuring product quality, yet it remains an open challenge owing to the extreme heterogeneity of defect morphologies—ranging from hairline cracks and microscopic pores to elongated scratches and shallow dents. Existing approaches, whether classical vision pipelines or recent deep-learning paradigms, struggle to simultaneously satisfy the stringent demands of industrial scenarios: high accuracy on sub-millimeter flaws, insensitivity to texture-rich backgrounds, and real-time throughput on resource-constrained hardware. Although contemporary detectors have narrowed the gap, they still exhibit pronounced sensitivity–robustness trade-offs, particularly in the presence of scale-varying defects and cluttered surfaces. To address these limitations, we introduce MBY (MBDNet-Attention-YOLO), a lightweight yet powerful framework that synergistically couples the MBDNet backbone with the YOLO detection head. Specifically, the backbone embeds three novel components: (1) HGStem, a hierarchical stem block that enriches low-level representations while suppressing redundant activations; (2) Dynamic Align Fusion (DAF), an adaptive cross-scale fusion mechanism that dynamically re-weights feature contributions according to defect saliency; and (3) C2f-DWR, a depth-wise residual variant that progressively expands receptive fields without incurring prohibitive computational costs. Building upon this enriched feature hierarchy, the neck employs our proposed MultiSEAM module—a cascaded squeeze-and-excitation attention mechanism operating at multiple granularities—to harmonize fine-grained and semantic cues, thereby amplifying weak defect signals against complex textures. Finally, we integrate the Inner-SIoU loss, which refines the geometric alignment between predicted and ground-truth boxes by jointly optimizing center distance, aspect ratio consistency, and IoU overlap, leading to faster convergence and tighter localization. Extensive experiments on two publicly available steel-defect benchmarks—NEU-DET and PVEL-AD—demonstrate the superiority of MBY. Without bells and whistles, our model achieves 85.8% mAP@0.5 on NEU-DET and 75.9% mAP@0.5 on PVEL-AD, surpassing the best-reported results by significant margins while maintaining real-time inference on an NVIDIA Jetson Xavier. Ablation studies corroborate the complementary roles of each component, underscoring MBY’s robustness across defect scales and surface conditions. These results suggest that MBY strikes an appealing balance between accuracy, efficiency, and deployability, offering a pragmatic solution for next-generation industrial quality-control systems. Full article

Accurate forecasting of tropical cyclone (TC) tracks is critical for disaster preparedness and risk mitigation. While traditional numerical weather prediction (NWP) systems have long served as the backbone of operational forecasting, they face limitations in computational cost and sensitivity to initial conditions. In recent years, deep learning (DL) has emerged as a promising alternative, offering data-driven modeling capabilities for capturing nonlinear spatiotemporal patterns. This paper presents a comprehensive review of DL-based approaches for TC track forecasting. We categorize all DL-based TC tracking models according to the architecture, including recurrent neural networks (RNNs), convolutional neural networks (CNNs), Transformers, graph neural networks (GNNs), generative models, and Fourier-based operators. To enable rigorous performance comparison, we introduce a Unified Geodesic Distance Error (UGDE) metric that standardizes evaluation across diverse studies and lead times. Based on this metric, we conduct a critical comparison of state-of-the-art models and identify key insights into their relative strengths, limitations, and suitable application scenarios. Building on this framework, we conduct a critical cross-model analysis that reveals key trends, performance disparities, and architectural tradeoffs. Our analysis also highlights several persistent challenges, such as long-term forecast degradation, limited physical integration, and generalization to extreme events, pointing toward future directions for developing more robust and operationally viable DL models for TC track forecasting. To support reproducibility and facilitate standardized evaluation, we release an open-source UGDE conversion tool on GitHub. Full article

In distance learning environments, learner engagement directly impacts attention, motivation, and academic performance. Signs of fatigue, negative affect, or critical remarks can warn of growing disengagement and potential dropout. However, most existing approaches rely on a single modality, visual or text-based, without providing a general view of learners’ cognitive and affective states. We propose a multimodal system that integrates three complementary analyzes: (1) a CNN-LSTM model augmented with warning signs such as PERCLOS and yawning frequency for fatigue detection, (2) facial emotion recognition by EmoNet and an LSTM to handle temporal dynamics, and (3) sentiment analysis of feedback by a fine-tuned BERT model. It was evaluated on three public benchmarks: DAiSEE for fatigue, AffectNet for emotion, and MOOC Review (Coursera) for sentiment analysis. The results show a precision of 88.5% for fatigue detection, 70% for emotion detection, and 91.5% for sentiment analysis. Aggregating these cues enables an accurate identification of disengagement periods and triggers individualized pedagogical interventions. These results, although based on independently sourced datasets, demonstrate the feasibility of an integrated approach to detecting disengagement and open the door to emotionally intelligent learning systems with potential for future work in real-time content personalization and adaptive learning assistance. Full article

Mangroves, as a key component of the blue-carbon ecosystem, have exceptional carbon sequestration capacity and are mainly distributed in tropical coastal regions. In the Solomon Islands, ongoing degradation of mangrove forests, primarily due to land conversion and timber exploitation, highlights an urgent need for high-resolution spatial data to inform effective conservation strategies. The present study introduces an efficient and accurate methodology for mapping mangrove habitats and prioritizing protection areas utilizing open-source satellite imagery and datasets available through the Google Earth Engine platform in conjunction with a U-Net deep learning algorithm. The model demonstrates high performance, achieving an F1-score of 0.834 and an overall accuracy of 0.96, in identifying mangrove distributions. The total mangrove area in the Solomon Islands is estimated to be approximately 71,348.27 hectares, accounting for about 2.47% of the national territory. Furthermore, based on the mapped mangrove habitats, an optimized hotspot analysis is performed to identify regions characterized by high-density mangrove distribution. By incorporating spatial variables such as distance from roads and urban centers, along with mangrove area, this study proposes priority mangrove protection areas. These results underscore the potential for using openly accessible satellite data to enhance the precision of mangrove conservation strategies in data-limited settings. This approach can effectively support coastal resource management and contribute to broader climate change mitigation strategies. Full article

Dual polarimetric SAR is capable of reflecting the biophysical and geometrical information of terrain with open access data availability. When it is combined with time-series observations, it can effectively capture the dynamic evolution of scattering characteristics of crops in different growth cycles. However, the actual planting of crops often shows spatial dispersion, and the same crop may be dispersed in different plots, which fails to adequately consider the correlation information between dispersed plots of the same crop in spatial distribution. This study proposed a crop classification method based on multi-temporal dual polarimetric data, which considered the utilization of information between near and far spatial plots, by employing superpixel segmentation and a HyperGraph neural network, respectively. Firstly, the method utilized the dual polarimetric covariance matrix of multi-temporal data to perform superpixel segmentation on neighboring pixels, so that the segmented superpixel blocks were highly compatible with the actual plot shapes from a long-term period perspective. Then, a HyperGraph adjacency matrix was constructed, and a HyperGraph neural network (HGNN) was utilized to better learn the features of plots of the same crop that are distributed far from each other. The method fully utilizes the three dimensions of time, polarization and space information, which complement each other so as to effectively realize high-precision crop classification. The Sentinel-1 experimental results show that, under the optimal parameter settings, the classified accuracy of combined temporal superpixel scattering features using the HGNN was obviously improved, considering the near and far distance spatial correlations of crop types. Full article

Automatic segmentation of nuclei on breast cancer histology images is a basic and important step for diagnosis in a computer-aided diagnostic approach and helps pathologists discover cancer early. Nuclei segmentation remains a challenging problem due to cancer biology and the variability of tissue characteristics; thus, their detection in an image is a very tedious and time-consuming task. In this context, overlapping nuclei objects present difficulties in separating them by conventional segmentation methods; thus, active contours can be employed in image segmentation. A major limitation of the active contours method is its inability to resolve image boundaries/edges of intersecting objects and segment multiple overlapping objects as a single object. Therefore, we present a hybrid active contour (connected component + active contours) method to segment cancerous lesions in unsupervised human breast histology images. Initially, this approach prepares and pre-processes data through various augmentation methods to increase the dataset size. Then, a stain normalization technique is applied to these augmented images to isolate nuclei features from tissue structures. Secondly, morphology operation techniques, namely erosion, dilation, opening, and distance transform, are used to highlight foreground and background pixels while removing overlapping regions from the highlighted nuclei objects on the image. Consequently, the connected components method groups these highlighted pixel components with similar intensity values and assigns them to their relevant labeled component to form a binary mask. Once all binary-masked groups have been determined, a deep-learning recurrent neural network (RNN) model from the Keras architecture uses this information to automatically segment nuclei objects having cancerous lesions on the image via the active contours method. This approach, therefore, uses the capabilities of connected components analysis to solve the limitations of the active contour method. This segmentation method is evaluated on an unsupervised, augmented human breast cancer histology dataset of 15,179 images. This proposed method produced a significant evaluation result of $98.71\%$ accuracy score. Full article

Access control systems protect homes and businesses in the continually evolving security industry. This paper designs and implements a Raspberry Pi-based facial recognition door lock system using artificial intelligence and computer vision for reliability, efficiency, and usability. With the Raspberry Pi as its CPU, the system uses facial recognition for authentication. A camera module for real-time image capturing, a relay module for solenoid lock control, and OpenCV for image processing are essential. The system uses the DeepFace library to detect user emotions and adaptive learning to improve recognition accuracy for approved users. The device also adapts to poor lighting and distances, and it sends real-time remote monitoring messages. Some of the most important things that have been achieved include adaptive facial recognition, ensuring that the system changes as it is used, and integrating real-time notifications and emotion detection without any problems. Face recognition worked well in many settings. Modular architecture facilitated hardware–software integration and scalability for various applications. In conclusion, this study created an intelligent facial recognition door lock system using Raspberry Pi hardware and open-source software libraries. The system addresses traditional access control limits and is practical, scalable, and inexpensive, demonstrating biometric technology’s potential in modern security systems. Full article

Background and Objectives: Depth perception of the human visual system in three-dimensional (3D) space plays an important role in human–computer interaction and artificial intelligence (AI) areas. It mainly employs binocular disparity and motion parallax cues. This study aims to systemically summarize the related studies about depth perception specified by these two cues. Materials and Methods: We conducted a literature investigation on related studies and summarized them from aspects like motivations, research trends, mechanisms, and interaction models of depth perception specified by these two cues. Results: Development trends show that depth perception research has gradually evolved from early studies based on a single cue to quantitative studies based on the interaction between these two cues. Mechanisms of these two cues reveal that depth perception specified by the binocular disparity cue is mainly influenced by factors like spatial variation in disparity, viewing distance, the position of visual field (or retinal image) used, and interaction with other cues; whereas that specified by the motion parallax cue is affected by head movement and retinal image motion, interaction with other cues, and the observer’s age. By integrating these two cues, several types of models for depth perception are summarized: the weak fusion (WF) model, the modified weak fusion (MWF) model, the strong fusion (SF) model, and the intrinsic constraint (IC) model. The merits and limitations of each model are analyzed and compared. Conclusions: Based on this review, a clear picture of the study on depth perception specified by binocular disparity and motion parallax cues can be seen. Open research challenges and future directions are presented. In the future, it is necessary to explore methods for easier manipulating of depth cue signals in stereoscopic images and adopting deep learning-related methods to construct models and predict depths, to meet the increasing demand of human–computer interaction in complex 3D scenarios. Full article

With the increasing mining depth, the stability of open-pit mine slopes has become an increasingly important concern. This study focuses on an open-pit mine in Southwest China and utilizes unmanned aerial vehicle (UAV) technology to gather data from these high and steep slopes. First, high-precision digital surface models and digital orthophoto data are collected using UAV terrain-following flight technology. However, two major challenges arise when applying geographic information systems (GISs) to this issue. The first challenge is that the extreme steepness of the slopes causes overlapping lithological layers at the same location, which GISs cannot resolve. The second challenge is that GISs cannot assess the influence of faults on landslides by calculating three-dimensional spatial distances. To overcome these issues, this study proposes the construction of a detailed 3D geological model for the entire mining area. This model allows for a more precise analysis of the lithology and fault spatial distances. A GIS is then applied to analyze the slope, curvature, and slope direction. Multi-source data fusion is employed to link spatial coordinates and create a dataset for further analysis. Five machine learning models for landslide prediction are compared using this dataset. Based on these comparisons, a high-precision random forest and slope boosting coupled method is developed to enhance the landslide prediction accuracy. Finally, a numerical simulation of a regional focus area is conducted, simulating the excavation process of an open-pit mine and analyzing the timing, location, and state of potential landslides. The results indicate that combining machine learning and multi-source data fusion provides a highly accurate, efficient, and straightforward method for landslide prediction in the high and steep slopes of open-pit mines. Full article

In rolling bearing fault diagnosis, when an unknown fault is present, the Closed-Set Recognition (CSR) method tends to misclassify it as a known fault. To address this issue, an Open-Set Recognition (OSR) framework is proposed for rolling bearing fault diagnosis in this study. The framework is built upon a serial multi-scale convolutional prototype learning (SMCPL) network, enhanced with an efficient channel attention (ECA) mechanism to extract the most critical fault features. The extracted features are fed into the Density Peak Clustering (DPC) module, which identifies known and unknown classes based on the computed local densities and relative distances. Finally, validation is performed on the Case Western Reserve University (CWRU) dataset, the Xi’an Jiaotong University rolling bearing accelerated life test dataset (XJTU-SY), and the Paderborn University bearing dataset (PU), Germany, and the framework is comprehensively evaluated in terms of several evaluation metrics, such as normalization accuracy and feature visualization. The experimental results show that SMCPL-ECA-DPC outperforms the comparative methods of SMCPL, CPL, ANEDL, CNN, and OpenMax and has high diagnostic performance in the identification of unknown faults. Full article

Accurately predicting the drift trajectory of green tides is crucial for assessing potential risks and implementing effective countermeasures. This paper proposes a short-term green-tide drift prediction method that combines green-tide patch characteristics, 1 h interval drift distances from GOCI-II images, and driving-factor data using the XGBoost machine learning model to enhance prediction accuracy. The results demonstrate that the proposed method outperforms the traditional OpenDrift model in short-term predictions. Specifically, at time intervals of 3, 5, and 7 h, the root mean square errors (RMSEs) of the OpenDrift model in the zonal direction are 1.81 km, 2.89 km, and 3.55 km, respectively, whereas the RMSEs of the proposed method are 0.80 km, 0.98 km, and 1.20 km, respectively; in the meridional direction, the RMSEs of the OpenDrift model are 1.77 km, 2.67 km, and 3.10 km, while the RMSEs for the proposed method are 0.82 km, 1.10 km, and 1.25 km, respectively. Furthermore, the proposed XGBoost method more-accurately tracks the actual positions of green-tide patches compared to the OpenDrift model. Specifically, at the 25 h interval, the proposed method continues to accurately predict patch positions, while the OpenDrift model exhibits significant deviations. This study demonstrates that the proposed method, by learning drift patterns from historical data, effectively predicts the short-term drift process of green tides. It provides valuable support for early warning systems, thereby helping to mitigate the ecological and economic impacts of green-tide disasters. Full article

Scientific site selection for urban parks is an important way to increase urban resilience and safeguard people’s well-being. Aiming at the lack of systematic consideration in the traditional park siting research, this study utilizes geographically weighted regression to explore the various characteristic factors affecting the spatial distribution of parks, and based on this, combines the random forest model and the interpretable model to accurately assess the potential of parks on urban land in Shenzhen and provide the basis for site selection. The study indicates that: ① Shenzhen’s parks exhibit complex differentiation characteristics in terms of natural landscape elements and the intensity of economic activities; ② The geographically weighted random forest (GWRF) model has better learning and generalization capabilities compared to the random forest (RF) model, and the average accuracy of the GWRF model is improved by 0.04 compared to the traditional RF model; ③ The park’s development potential is divided according to the results of the GWRF model, with 52.01% denoted as the potential incubation zone, 21.15% the potential accumulation zone, 8.25% the potential growth zone, and 18.59% the potential core zone; ④ Through interpretability analysis, it is identified that vegetation coverage, the density of tourist attractions or points of interest (POI), slope, elevation, and nighttime light intensity are the most significant factors affecting park development potential, while the distance to roads and the distance to bodies of water are the least influential factors. The research systematically explores a quantitative evaluation framework for the development potential of Shenzhen’s parks, opening new theoretical pathways and practical paradigms for the sustainable development planning of Shenzhen under the “Park City” concept. Full article

The advancement of SAR satellites enables continuous and real-time ship monitoring on water surfaces regardless of time and weather. Traditional ship detection algorithms in SAR imagery using manually designed operators lack accuracy, while many existing deep learning-based detection algorithms are computationally intensive and have room for accuracy improvement. Inspired by CenterNet, we propose the Three Points Network (TPNet). It locates the ship’s center point and estimates distances to the top-left and bottom-right corners for precise positioning. We introduce several innovative mechanisms to enhance TPNet’s performance, improving both accuracy and computational efficiency. Evaluated on the open-source SAR-Ship-Dataset, TPNet outperforms 14 other deep learning-based detection algorithms in accuracy and efficiency. Its strong generalization ability is further verified on SSDD and HRSID datasets. These results show TPNet’s potential in real-time maritime surveillance and monitoring systems. Full article

A new machine learning algorithm, BoxRF, was developed specifically for estimating grades from drillhole datasets. The method combines the features of classical estimation methods, such as search boxes, search direction, and estimation based on inverse distance methods, with the robustness of random forest (RF) methods that come from forming numerous random groups of data. The method was applied to a porphyry copper deposit, and results were compared to various ML methods, including XGBoost (XGB), k-nearest neighbors (KNN), neural nets (NN), and RF. Scikit-learn RF (SRF) performed the best (R² = 0.696) among the ML methods but underperformed BoxRF (R² = 0.751). The results were confirmed through a five-fold cross-validation exercise where BoxRF once again outperformed SRF. The box dimensions that performed the best were similar in length to the ranges indicated by variogram modeling, thus demonstrating a link between machine learning and traditional methods. Numerous combinations of hyperparameters performed similarly well, implying the method is robust. The inverse distance method was found to better represent the grade–space relationship in BoxRF than median values. The superiority of BoxRF over SRF in this dataset is encouraging, as it opens the possibility of improving machine learning by incorporating domain knowledge (principles of geology, in this case). Full article

: A substantial amount of research has demonstrated the robustness and accuracy of the Riemannian minimum distance to mean (MDM) classifier for all kinds of EEG-based brain–computer interfaces (BCIs). This classifier is simple, fully deterministic, robust to noise, computationally efficient, and prone to transfer learning. Its training is very simple, requiring just the computation of a geometric mean of a symmetric positive-definite (SPD) matrix per class. We propose an improvement of the MDM involving a number of power means of SPD matrices instead of the sole geometric mean. By the analysis of 20 public databases, 10 for the motor-imagery BCI paradigm and 10 for the P300 BCI paradigm, comprising 587 individuals in total, we show that the proposed classifier clearly outperforms the MDM, approaching the state-of-the art in terms of performance while retaining the simplicity and the deterministic behavior. In order to promote reproducible research, our code will be released as open source. Full article