Search Results (2,567)

This study proposes Dual-Stream Former, a novel architecture that integrates a Video Swin Transformer and Conformer designed to address the challenges of visual speech recognition (VSR). The model captures spatiotemporal dependencies, achieving a state-of-the-art character error rate (CER) of 3.46%, surpassing traditional convolutional neural network (CNN)-based models, such as 3D-CNN + DenseNet-121 (CER: 5.31%), and transformer-based alternatives, such as vision transformers (CER: 4.05%). The Video Swin Transformer captures multiscale spatial representations with high computational efficiency, whereas the Conformer back-end enhances temporal modeling across diverse phoneme categories. Evaluation of a high-resolution dataset comprising 740,000 utterances across 185 classes highlighted the effectiveness of the model in addressing visually confusing phonemes, such as diphthongs (/ai/, /au/) and labio-dental sounds (/f/, /v/). Dual-Stream Former achieved phoneme recognition error rates of 10.39% for diphthongs and 9.25% for labiodental sounds, surpassing those of CNN-based architectures by more than 6%. Although the model’s large parameter count (168.6 M) poses resource challenges, its hierarchical design ensures scalability. Future work will explore lightweight adaptations and multimodal extensions to increase deployment feasibility. These findings underscore the transformative potential of Dual-Stream Former for advancing VSR applications such as silent communication and assistive technologies by achieving unparalleled precision and robustness in diverse settings. Full article

Asphalt crack segmentation is essential for preventive maintenance but is sensitive to noise, viewpoint, and illumination. This study evaluates a minimally invasive strategy that augments standard RGB input with an auxiliary fourth channel—a crack-probability map generated by a multi-scale ensemble of classifiers—and injects it into segmentation backbones. Field imagery from unmanned aerial vehicles and action cameras was used to train and compare U-Net, ENet, HRNet, and DeepLabV3+ under unified settings; the probability map was produced by an ensemble of lightweight convolutional neural networks (CNNs). Across models, the four-channel configuration improved performance over three-channel baselines; for DeepLabV3+, the Intersection over Union (IoU) increased by 6.41%. Transformer-based classifiers, despite strong accuracy, proved less effective and slower than lightweight CNNs for probability-map generation; the final ensemble processed images in approximately 0.63 s each. Integrating ensemble-derived probability maps yielded consistent gains, with the best four-channel CNNs surpassing YOLO11x-seg and Transformer baselines while remaining practical. This study presents a systematic evaluation showing that probability maps from classifier ensembles can serve as an auxiliary channel to improve segmentation of asphalt pavement cracks, providing a novel modular complement or alternative to attention mechanisms. The findings demonstrate a practical and effective strategy for enhancing automated pavement monitoring. Full article

Bridge Structural Health Monitoring (BSHM) is vital for assessing structural integrity and operational safety. Traditional wired systems are limited by high installation costs and complexity, while existing wireless systems still face issues with cost, synchronization, and reliability. Moreover, cloud-based methods for extreme event detection struggle to meet real-time and bandwidth constraints in edge environments. To address these challenges, this study proposes a lightweight wireless BSHM system based on edge computing, enabling local data acquisition and real-time intelligent detection of extreme events. The system consists of wireless sensor nodes for front-end acceleration data collection and an intelligent hub for data storage, visualization, and earthquake recognition. Acceleration data are converted into time–frequency images to train a MobileNetV2-based model. With model quantization and Neural Processing Unit (NPU) acceleration, efficient on-device inference is achieved. Experiments on a laboratory steel bridge verify the system’s high acquisition accuracy, precise clock synchronization, and strong anti-interference performance. Compared with inference on a general-purpose ARM CPU running the unquantized model, the quantized model deployed on the NPU achieves a 26× speedup in inference, a 35% reduction in power consumption, and less than 1% accuracy loss. This solution provides a cost-effective, reliable BSHM framework for small-to-medium-sized bridges, offering local intelligence and rapid response with strong potential for real-world applications. Full article

To tackle the challenges of complex construction environment interference (e.g., lighting variations, occlusion, and marker contamination) and the demand for high-precision alignment during the hoisting process of bridge-erecting machines, this paper presents a two-stage marker detection–localization network tailored to hoisting alignment. The proposed network adopts a “coarse detection–fine estimation” phased framework; the first stage employs a lightweight detection module, which integrates a dynamic hybrid backbone (DHB) and dynamic switching mechanism to efficiently filter background noise and generate coarse localization boxes of marker regions. Specifically, the DHB dynamically switches between convolutional and Transformer branches to handle features of varying complexity (using depthwise separable convolutions from MobileNetV3 for low-level geometric features and lightweight Transformer blocks for high-level semantic features). The second stage constructs a Transformer-based homography estimation module, which leverages multi-head self-attention to capture long-range dependencies between marker keypoints and the scene context. By integrating enhanced multi-scale feature interaction and position encoding (combining the absolute position and marker geometric priors), this module achieves the end-to-end learning of precise homography matrices between markers and hoisting equipment from the coarse localization boxes. To address data scarcity in construction scenes, a multi-dimensional data augmentation strategy is developed, including random homography transformation (simulating viewpoint changes), photometric augmentation (adjusting brightness, saturation, and contrast), and background blending with bounding box extraction. Experiments on a real bridge-erecting machine dataset demonstrate that the network achieves detection accuracy (mAP) of 97.8%, a homography estimation reprojection error of less than 1.2 mm, and a processing frame rate of 32 FPS. Compared with traditional single-stage CNN-based methods, it significantly improves the alignment precision and robustness in complex environments, offering reliable technical support for the precise control of automated hoisting in bridge-erecting machines. Full article

Breast cancer is one of the leading causes of death among women of all ages and backgrounds globally. In recent years, the growing deficit of expert radiologists—particularly in underdeveloped countries—alongside a surge in the number of images for analysis, has negatively affected the ability to secure timely and precise diagnostic results in breast cancer screening. AI technologies offer powerful tools that allow for the effective diagnosis and survival forecasting, reducing the dependency on human cognitive input. Towards this aim, this research introduces a deep meta-learning framework for swift analysis of mammography images—combining a Siamese network model with a triplet-based loss function—to facilitate automatic screening (recognition) of potentially suspicious breast cancer cases. Three pre-trained deep CNN architectures, namely GoogLeNet, ResNet50, and MobileNetV3, are fine-tuned and scrutinized for their effectiveness in transforming input mammograms to a suitable embedding space. The proposed framework undergoes a comprehensive evaluation through a rigorous series of experiments, utilizing two different, publicly accessible, and widely used datasets of digital X-ray mammograms: INbreast and CBIS-DDSM. The experimental results demonstrate the framework’s strong performance in differentiating between tumorous and normal images, even with a very limited number of training samples, on both datasets. Full article

Addressing the carbon footprint in pig production is a fundamental technical basis for achieving carbon neutrality and peak carbon emissions. Only by systematically studying the carbon footprint can the goals of carbon neutrality and peak carbon emissions be effectively realized. This study aims to reduce the carbon footprint through optimized feeding strategies based on minimizing carbon emissions. To this end, this study conducted a full-lifecycle monitoring of the carbon footprint during pig growth from December 2024 to May 2025, optimizing feeding strategies using a real-time pig weight estimation model driven by deep learning to reduce resource consumption and the carbon footprint. We introduce EcoSegLite, a lightweight deep learning model designed for non-contact real-time pig weight estimation. By incorporating ShuffleNetV2, Linear Deformable Convolution (LDConv), and ACmix modules, it achieves high precision in resource-constrained environments with only 1.6 M parameters, attaining a 96.7% mAP50. Based on full-lifecycle weight monitoring of 63 pigs at the Pianguan farm from December 2024 to May 2025, the EcoSegLite model was integrated with a life cycle assessment (LCA) framework to optimize feeding management. This approach achieved a 7.8% reduction in feed intake, an 11.9% reduction in manure output, and a 5.1% reduction in carbon footprint. The resulting growth curves further validated the effectiveness of the optimized feeding strategy, while the reduction in feed and manure also potentially reduced water consumption and nitrogen runoff. This study offers a data-driven solution that enhances resource efficiency and reduces environmental impact, paving new pathways for precision agriculture and sustainable livestock production. Full article

Texture, a significant visual attribute in images, plays an important role in many pattern recognition tasks. While Convolutional Neural Networks (CNNs) have been among the most effective methods for texture analysis, alternative architectures such as Vision Transformers (ViTs) have recently demonstrated superior performance on a range of visual recognition problems. However, the suitability of ViTs for texture recognition remains underexplored. In this work, we investigate the capabilities and limitations of ViTs for texture recognition by analyzing 25 different ViT variants as feature extractors and comparing them to CNN-based and hand-engineered approaches. Our evaluation encompasses both accuracy and efficiency, aiming to assess the trade-offs involved in applying ViTs to texture analysis. Our results indicate that ViTs generally outperform CNN-based and hand-engineered models, particularly when using strong pre-training and in-the-wild texture datasets. Notably, BeiTv2-B/16 achieves the highest average accuracy (85.7%), followed by ViT-B/16-DINO (84.1%) and Swin-B (80.8%), outperforming the ResNet50 baseline (75.5%) and the hand-engineered baseline (73.4%). As a lightweight alternative, EfficientFormer-L3 attains a competitive average accuracy of 78.9%. In terms of efficiency, although ViT-B and BeiT(v2) have a higher number of GFLOPs and parameters, they achieve significantly faster feature extraction on GPUs compared to ResNet50. These findings highlight the potential of ViTs as a powerful tool for texture analysis while also pointing to areas for future exploration, such as efficiency improvements and domain-specific adaptations. Full article

Coniferous forests in Canada play a vital role in carbon sequestration, wildlife conservation, climate change mitigation, and long-term sustainability. Traditional methods for classifying and segmenting coniferous trees have primarily relied on the direct use of spectral or LiDAR-based data. In 2024, we introduced a novel data representation method, pseudo tree crown (PTC), which provides a pseudo-3D pixel-value view that enhances the informational richness of images and significantly improves classification performance. While our original implementation was successfully tested on urban and deciduous trees, this study extends the application of PTC to Canadian conifer species, including jack pine, Douglas fir, spruce, and aspen. We address key challenges such as snow-covered backgrounds and evaluate the impact of training dataset size on classification results. Classification was performed using Random Forest, PyTorch (ResNet50), and YOLO versions v10, v11, and v12. The results demonstrate that PTC can substantially improve individual tree classification accuracy by up to 13%, reaching the high 90% range. Full article

To address the high fossil energy dependency and the low-value utilization of stillage (WDGS) in conventional cassava-based ethanol production—factors that increase greenhouse gas emissions and limit overall sustainability—this study develops an integrated ethanol–biogas–CHP system that valorizes stillage and enhances energy recovery. Three process scenarios were designed and evaluated through life cycle assessment (LCA) and techno-economic analysis: Case-I (WDGS dried and sold as animal feed), Case-II (stillage anaerobically digested for biogas used for heat), and Case-III (biogas further utilized in a combined heat and power system). Process simulation was conducted in Aspen Plus V11, while environmental impacts were quantified with the CML 2001 methodology under a cradle-to-gate boundary across six categories, including global warming potential (GWP) and abiotic depletion potential (ADP). Results show that Case-III achieves the highest environmental and economic performance, with a net GWP of −1515.05 kg CO₂-eq/ton ethanol and the greatest profit of 396.80 USD/ton of ethanol, attributed to internal energy self-sufficiency and surplus electricity generation. Sensitivity analysis further confirms Case-III’s robustness under variations in transportation distance and electricity demand. Overall, valorizing cassava stillage through biogas–CHP integration significantly improves the sustainability of ethanol production, offering a practical pathway toward low-carbon bioenergy with potential for negative emissions. This study fills a gap in previous life cycle research by jointly assessing WDGS utilization pathways with techno-economic evaluation, providing actionable insights for carbon-neutral bioenergy policies in cassava-producing regions. Certain limitations, such as software version and data accessibility, remain to be addressed in future work. Full article

Valorisation of spent yeast biomass post-fermentation requires energy-intensive autolysis or enzymatic hydrolysis that reduces the net benefit. Here, we present a simple and reproducible method for generating functional yeast extract recycled from Komagataella phaffii biomass without a requirement of a pre-treatment process. Spent yeast pellets from fermentations were freeze-dried to produce a fine powder that can be used directly at low concentrations, 0.0015% (w/v), together with 2% peptone (w/v), to formulate complete media ready for secondary fermentations. This media formulation supported growth rates of yeast culture that were statistically indistinguishable (p-value > 0.05) from cultures grown in standard YPD media containing commercial yeast extract, and these cultures produced equivalent titres of recombinant β-glucosidase (0.998 Abs_405nm commercial extract vs. 0.899 Abs_405nm recycled extract). Additionally, nutrient analyses highlight equivalent levels of sugars (~23 g/L), total proteins, and cell yield per carbon source (~2.17 g) with this recycled yeast extract media formulation when compared to commercial media. This method reduces process complexity and cost and enables the circular reuse of yeast biomass. The protocol is technically straightforward to implement, using freeze drying that is commonly available in research laboratories, representing a broadly applicable and sustainable alternative to conventional media supplementation that achieves a circular approach within the same fermentation system. Full article

Bats are ecologically vital mammals, serving as pollinators, seed dispersers, and bioindicators of ecosystem health. Many species inhabit natural caves, which offer optimal conditions for survival but present challenges for direct ecological monitoring due to their dark, complex, and inaccessible environments. Traditional monitoring methods, such as mist-netting, are invasive and limited in scope, highlighting the need for non-intrusive alternatives. In this work, we present a portable multisensor platform designed to operate in underground habitats. The system captures multimodal data, including near-infrared (NIR) imagery, ultrasonic audio, 3D structural data, and RGB video. Focusing on NIR imagery, we evaluate the effectiveness of the YOLO object detection framework for automated bat detection and counting. Experiments were conducted using a dataset of NIR images collected in natural shelters. Three YOLO variants (v10, v11, and v12) were trained and tested on this dataset. The models achieved high detection accuracy, with YOLO v12m reaching a mean average precision (mAP) of 0.981. These results demonstrate that combining NIR imaging with deep learning enables accurate and non-invasive monitoring of bats in challenging environments. The proposed approach offers a scalable tool for ecological research and conservation, supporting population assessment and behavioral studies without disturbing bat colonies. Full article

Diagnosing gastrointestinal disorders (GIDs) remains a significant challenge, particularly when relying on wireless capsule endoscopy (WCE), which lacks advanced imaging enhancements like Narrow Band Imaging (NBI). To address this, we propose a novel framework, the Spectrum-Aided Vision Enhancer (SAVE), especially designed to transform standard white light (WLI) endoscopic images into spectrally enriched representations that emulate both hyperspectral imaging (HSI) and NBI formats. By leveraging color calibration through the Macbeth Color Checker, gamma correction, CIE 1931 XYZ transformation, and principal component analysis (PCA), SAVE reconstructs detailed spectral information from conventional RGB inputs. Performance was evaluated using the Kvasir-v2 dataset, which includes 6490 annotated images spanning eight GI-related categories. Deep learning models like Inception-Net V3, MobileNetV2, MobileNetV3, and AlexNet were trained on both original WLI- and SAVE-enhanced images. Among these, MobileNetV2 achieved an F1-score of 96% for polyp classification using SAVE, and AlexNet saw a notable increase in average accuracy to 84% when applied to enhanced images. Image quality assessment showed high structural similarity (SSIM scores of 93.99% for Olympus endoscopy and 90.68% for WCE), confirming the fidelity of the spectral transformations. Overall, the SAVE framework offers a practical, software-based enhancement strategy that significantly improves diagnostic accuracy in GI imaging, with strong implications for low-cost, non-invasive diagnostics using capsule endoscopy systems. Full article

This study evaluates the financial viability of different main engine–fuel configurations for a container vessel on a standardized Trans-Pacific route. Using Net Present Value (NPV) analysis over a 10 year evaluation period (2024–2033), it compares six propulsion scenarios, combining three Wärtsilä engine types and four fuel alternatives (HFO, LFO, LNG, Methanol). The framework integrates technical parameters, including engine power and fuel consumption, with financial instruments such as the Weighted Average Cost of Capital (WACC) and the Capital Asset Pricing Model (CAPM). Results show that the LNG-powered Wärtsilä 8V31DF achieves the highest NPV. Despite requiring the highest initial capital expenditure (CAPEX), this configuration delivers superior financial performance and remains robust under volatile market conditions. Sensitivity tests with ±20% freight–fuel shocks and alternative discount rates (5%, 7.18%, 10%) confirm that the relative ranking of propulsion options is stable. Methanol yields negative NPVs under current prices but could become competitive with bio-methanol cost reductions or strong carbon pricing. Limitations include constant non-fuel OPEX, fixed sea state, and the exclusion of explicit carbon price scenarios. From a policy perspective, LNG appears most viable in the short term, while long-term strategies should consider ammonia and hydrogen in line with IMO decarbonization pathways. Full article

Background/Objectives: Vertical growth pattern significantly influences facial aesthetics and treatment choices. Lateral cephalograms are routinely used for the evaluation of vertical jaw relationships in orthodontic diagnosis. The aim of this study was to evaluate the performance of deep learning algorithms in classifying cephalometric radiographs according to vertical skeletal growth patterns without the need for anatomical landmark identification. Methods: This study was carried out on lateral cephalometric radiographs of 1050 patients. Cephalometric radiographs were divided into 3 subgroups based on FMA, SN-GoGn, and Cant of Occlusal Plane angles. Six deep learning models (ResNet101, DenseNet 201, EfficientNet B0, EfficientNet V2 B0, ConvNetBase, and a hybrid model) were employed for the classification of the dataset. The performances of the well-known deep learning models and the hybrid model were compared for accuracy, precision, F1-Score, mean absolute error, Cohen’s Kappa, and Grad-CAM metrics. Results: The highest accuracy rates were achieved by the Hybrid Model with 86.67% for FMA groups, 87.29% for SN-GoGn groups, and 82.71% for Cant of Occlusal Plane groups. The lowest accuracy rates were achieved by ConvNet with 79.58% for FMA groups, 65% for SN-GoGn, and 70.21% for Cant of Occlusal Plane groups. Conclusions: The six deep learning algorithms employed demonstrated classification success rates ranging from 65% to 87.29%. The highest classification accuracy was observed in the FMA angle, while the lowest accuracy was recorded for the Cant of the Occlusal Plane angle. The proposed DL algorithms showed potential for direct skeletal orthodontic diagnosis without the need for cephalometric landmark detection steps. Full article

The traditional detection of blueberry maturity relies on human experience, which is inefficient and highly subjective. Although deep learning methods have improved accuracy, they require large models and complex computations, making real-time deployment on resource-constrained edge devices difficult. To address these issues, a GLL-YOLO method based on the YOLOv8 network is proposed to deal with problems such as fruit occlusion and complex backgrounds in mature blueberry detection. This approach utilizes the GhostNetV2 network as the backbone. The LIMC module is suggested to substitute the original C2f module. Meanwhile, a Lightweight Shared Convolution Detection Head (LSCD) module is designed to build the GLL-YOLO model. This model can accurately detect blueberries at three different maturity stages: unripe, semi-ripe, and ripe. It significantly reduces the number of model parameters and floating-point operations while maintaining high accuracy. Experimental results show that GLL-YOLO outperforms the original YOLOv8 model in terms of accuracy, with mAP improvements of 4.29%, 1.67%, and 1.39% for unripe, semi-ripe, and ripe blueberries, reaching 94.51%, 91.72%, and 93.32%, respectively. Compared to the original model, GLL-YOLO improved the accuracy, recall rate, and mAP by 2.3%, 5.9%, and 1%, respectively. Meanwhile, GLL-YOLO reduces parameters, FLOPs, and model size by 50%, 39%, and 46.7%, respectively, while maintaining accuracy. This method has the advantages of a small model size, high accuracy, and good detection performance, providing reliable support for intelligent blueberry harvesting. Full article