Modern manufacturing processes have become more reliant on automation because of the accelerated transition from Industry 3.0 to Industry 4.0.Manual inspection of products on assembly lines remains inefficient,prone t...Modern manufacturing processes have become more reliant on automation because of the accelerated transition from Industry 3.0 to Industry 4.0.Manual inspection of products on assembly lines remains inefficient,prone to errors and lacks consistency,emphasizing the need for a reliable and automated inspection system.Leveraging both object detection and image segmentation approaches,this research proposes a vision-based solution for the detection of various kinds of tools in the toolkit using deep learning(DL)models.Two Intel RealSense D455f depth cameras were arranged in a top down configuration to capture both RGB and depth images of the toolkits.After applying multiple constraints and enhancing them through preprocessing and augmentation,a dataset consisting of 3300 annotated RGB-D photos was generated.Several DL models were selected through a comprehensive assessment of mean Average Precision(mAP),precision-recall equilibrium,inference latency(target≥30 FPS),and computational burden,resulting in a preference for YOLO and Region-based Convolutional Neural Networks(R-CNN)variants over ViT-based models due to the latter’s increased latency and resource requirements.YOLOV5,YOLOV8,YOLOV11,Faster R-CNN,and Mask R-CNN were trained on the annotated dataset and evaluated using key performance metrics(Recall,Accuracy,F1-score,and Precision).YOLOV11 demonstrated balanced excellence with 93.0%precision,89.9%recall,and a 90.6%F1-score in object detection,as well as 96.9%precision,95.3%recall,and a 96.5%F1-score in instance segmentation with an average inference time of 25 ms per frame(≈40 FPS),demonstrating real-time performance.Leveraging these results,a YOLOV11-based windows application was successfully deployed in a real-time assembly line environment,where it accurately processed live video streams to detect and segment tools within toolkits,demonstrating its practical effectiveness in industrial automation.The application is capable of precisely measuring socket dimensions by utilising edge detection techniques on YOLOv11 segmentation masks,in addition to detection and segmentation.This makes it possible to do specification-level quality control right on the assembly line,which improves the ability to examine things in real time.The implementation is a big step forward for intelligent manufacturing in the Industry 4.0 paradigm.It provides a scalable,efficient,and accurate way to do automated inspection and dimensional verification activities.展开更多
Microscopy imaging is fundamental in analyzing bacterial morphology and dynamics,offering critical insights into bacterial physiology and pathogenicity.Image segmentation techniques enable quantitative analysis of bac...Microscopy imaging is fundamental in analyzing bacterial morphology and dynamics,offering critical insights into bacterial physiology and pathogenicity.Image segmentation techniques enable quantitative analysis of bacterial structures,facilitating precise measurement of morphological variations and population behaviors at single-cell resolution.This paper reviews advancements in bacterial image segmentation,emphasizing the shift from traditional thresholding and watershed methods to deep learning-driven approaches.Convolutional neural networks(CNNs),U-Net architectures,and three-dimensional(3D)frameworks excel at segmenting dense biofilms and resolving antibiotic-induced morphological changes.These methods combine automated feature extraction with physics-informed postprocessing.Despite progress,challenges persist in computational efficiency,cross-species generalizability,and integration with multimodal experimental workflows.Future progress will depend on improving model robustness across species and imaging modalities,integrating multimodal data for phenotype-function mapping,and developing standard pipelines that link computational tools with clinical diagnostics.These innovations will expand microbial phenotyping beyond structural analysis,enabling deeper insights into bacterial physiology and ecological interactions.展开更多
This systematic review aims to comprehensively examine and compare deep learning methods for brain tumor segmentation and classification using MRI and other imaging modalities,focusing on recent trends from 2022 to 20...This systematic review aims to comprehensively examine and compare deep learning methods for brain tumor segmentation and classification using MRI and other imaging modalities,focusing on recent trends from 2022 to 2025.The primary objective is to evaluate methodological advancements,model performance,dataset usage,and existing challenges in developing clinically robust AI systems.We included peer-reviewed journal articles and highimpact conference papers published between 2022 and 2025,written in English,that proposed or evaluated deep learning methods for brain tumor segmentation and/or classification.Excluded were non-open-access publications,books,and non-English articles.A structured search was conducted across Scopus,Google Scholar,Wiley,and Taylor&Francis,with the last search performed in August 2025.Risk of bias was not formally quantified but considered during full-text screening based on dataset diversity,validation methods,and availability of performance metrics.We used narrative synthesis and tabular benchmarking to compare performance metrics(e.g.,accuracy,Dice score)across model types(CNN,Transformer,Hybrid),imaging modalities,and datasets.A total of 49 studies were included(43 journal articles and 6 conference papers).These studies spanned over 9 public datasets(e.g.,BraTS,Figshare,REMBRANDT,MOLAB)and utilized a range of imaging modalities,predominantly MRI.Hybrid models,especially ResViT and UNetFormer,consistently achieved high performance,with classification accuracy exceeding 98%and segmentation Dice scores above 0.90 across multiple studies.Transformers and hybrid architectures showed increasing adoption post2023.Many studies lacked external validation and were evaluated only on a few benchmark datasets,raising concerns about generalizability and dataset bias.Few studies addressed clinical interpretability or uncertainty quantification.Despite promising results,particularly for hybrid deep learning models,widespread clinical adoption remains limited due to lack of validation,interpretability concerns,and real-world deployment barriers.展开更多
Deep learning-based methods have become alternatives to traditional numerical weather prediction systems,offering faster computation and the ability to utilize large historical datasets.However,the application of deep...Deep learning-based methods have become alternatives to traditional numerical weather prediction systems,offering faster computation and the ability to utilize large historical datasets.However,the application of deep learning to medium-range regional weather forecasting with limited data remains a significant challenge.In this work,three key solutions are proposed:(1)motivated by the need to improve model performance in data-scarce regional forecasting scenarios,the authors innovatively apply semantic segmentation models,to better capture spatiotemporal features and improve prediction accuracy;(2)recognizing the challenge of overfitting and the inability of traditional noise-based data augmentation methods to effectively enhance model robustness,a novel learnable Gaussian noise mechanism is introduced that allows the model to adaptively optimize perturbations for different locations,ensuring more effective learning;and(3)to address the issue of error accumulation in autoregressive prediction,as well as the challenge of learning difficulty and the lack of intermediate data utilization in one-shot prediction,the authors propose a cascade prediction approach that effectively resolves these problems while significantly improving model forecasting performance.The method achieves a competitive result in The East China Regional AI Medium Range Weather Forecasting Competition.Ablation experiments further validate the effectiveness of each component,highlighting their contributions to enhancing prediction performance.展开更多
Recent advances in deep learning have significantly improved flood detection and segmentation from aerial and satellite imagery.However,conventional convolutional neural networks(CNNs)often struggle in complex flood s...Recent advances in deep learning have significantly improved flood detection and segmentation from aerial and satellite imagery.However,conventional convolutional neural networks(CNNs)often struggle in complex flood scenarios involving reflections,occlusions,or indistinct boundaries due to limited contextual modeling.To address these challenges,we propose a hybrid flood segmentation framework that integrates a Vision Transformer(ViT)encoder with a U-Net decoder,enhanced by a novel Flood-Aware Refinement Block(FARB).The FARB module improves boundary delineation and suppresses noise by combining residual smoothing with spatial-channel attention mechanisms.We evaluate our model on a UAV-acquired flood imagery dataset,demonstrating that the proposed ViTUNet+FARB architecture outperforms existing CNN and Transformer-based models in terms of accuracy and mean Intersection over Union(mIoU).Detailed ablation studies further validate the contribution of each component,confirming that the FARB design significantly enhances segmentation quality.To its better performance and computational efficiency,the proposed framework is well-suited for flood monitoring and disaster response applications,particularly in resource-constrained environments.展开更多
The accurate segmentation of deep gray matter nuclei is critical for neuropathological research,disease diagnosis and treatment.Existing methods employ the supervised learning training approach,which requires large la...The accurate segmentation of deep gray matter nuclei is critical for neuropathological research,disease diagnosis and treatment.Existing methods employ the supervised learning training approach,which requires large labeled datasets.It is challenging and time-consuming to obtain such datasets for medical image analysis.In addition,these methods based on convolutional neural networks(CNNs)only achieve suboptimal performance due to the locality of convolutional operations.Vision Transformers(ViTs)efficiently model long-range dependencies and thus have the potentiality to outperform these methods in segmentation tasks.To address these issues,we propose a novel hybrid network based on self-supervised pre-training for deep gray matter nuclei segmentation.Specifically,we present a CNN-Transformer hybrid network(CTNet),whose encoder consists of 3D CNN and ViT to learn local spatial-detailed features and global semantic information.A self-supervised learning(SSL)approach that integrates rotation prediction and masked feature reconstruction is proposed to pre-train the CTNet,enabling the model to learn valuable visual representations from unlabeled data.We evaluate the effectiveness of our method on 3T and 7T human brain MRI datasets.The results demonstrate that our CTNet achieves better performance than other comparison models and our pre-training strategy outperforms other advanced self-supervised methods.When the training set has only one sample,our pre-trained CTNet enhances segmentation performance,showing an 8.4%improvement in Dice similarity coefficient(DSC)compared to the randomly initialized CTNet.展开更多
Recent studies indicate that millions of individuals suffer from renal diseases,with renal carcinoma,a type of kidney cancer,emerging as both a chronic illness and a significant cause of mortality.Magnetic Resonance I...Recent studies indicate that millions of individuals suffer from renal diseases,with renal carcinoma,a type of kidney cancer,emerging as both a chronic illness and a significant cause of mortality.Magnetic Resonance Imaging(MRI)and Computed Tomography(CT)have become essential tools for diagnosing and assessing kidney disorders.However,accurate analysis of thesemedical images is critical for detecting and evaluating tumor severity.This study introduces an integrated hybrid framework that combines three complementary deep learning models for kidney tumor segmentation from MRI images.The proposed framework fuses a customized U-Net and Mask R-CNN using a weighted scheme to achieve semantic and instance-level segmentation.The fused outputs are further refined through edge detection using Stochastic FeatureMapping Neural Networks(SFMNN),while volumetric consistency is ensured through Improved Mini-Batch K-Means(IMBKM)clustering integrated with an Encoder-Decoder Convolutional Neural Network(EDCNN).The outputs of these three stages are combined through a weighted fusion mechanism,with optimal weights determined empirically.Experiments on MRI scans from the TCGA-KIRC dataset demonstrate that the proposed hybrid framework significantly outperforms standalone models,achieving a Dice Score of 92.5%,an IoU of 87.8%,a Precision of 93.1%,a Recall of 90.8%,and a Hausdorff Distance of 2.8 mm.These findings validate that the weighted integration of complementary architectures effectively overcomes key limitations in kidney tumor segmentation,leading to improved diagnostic accuracy and robustness in medical image analysis.展开更多
The deep learning technology has shown impressive performance in various vision tasks such as image classification, object detection and semantic segmentation. In particular, recent advances of deep learning technique...The deep learning technology has shown impressive performance in various vision tasks such as image classification, object detection and semantic segmentation. In particular, recent advances of deep learning techniques bring encouraging performance to fine-grained image classification which aims to distinguish subordinate-level categories, such as bird species or dog breeds. This task is extremely challenging due to high intra-class and low inter-class variance. In this paper, we review four types of deep learning based fine-grained image classification approaches, including the general convolutional neural networks （CNNs）, part detection based, ensemble of networks based and visual attention based fine-grained image classification approaches. Besides, the deep learning based semantic segmentation approaches are also covered in this paper. The region proposal based and fully convolutional networks based approaches for semantic segmentation are introduced respectively.展开更多
Microstructural classification is typically done manually by human experts,which gives rise to uncertainties due to subjectivity and reduces the overall efficiency.A high-throughput characterization is proposed based ...Microstructural classification is typically done manually by human experts,which gives rise to uncertainties due to subjectivity and reduces the overall efficiency.A high-throughput characterization is proposed based on deep learning,rapid acquisition technology,and mathematical statistics for the recognition,segmentation,and quantification of microstructure in weathering steel.The segmentation results showed that this method was accurate and efficient,and the segmentation of inclusions and pearlite phase achieved accuracy of 89.95%and 90.86%,respectively.The time required for batch processing by MIPAR software involving thresholding segmentation,morphological processing,and small area deletion was 1.05 s for a single image.By comparison,our system required only 0.102 s,which is ten times faster than the commercial software.The quantification results were extracted from large volumes of sequential image data(150 mm^(2),62,216 images,1024×1024 pixels),which ensure comprehensive statistics.Microstructure information,such as three-dimensional density distribution and the frequency of the minimum spatial distance of inclusions on the sample surface of 150 mm^(2),were quantified by extracting the coordinates and sizes of individual features.A refined characterization method for two-dimensional structures and spatial information that is unattainable when performing manually or with software is provided.That will be useful for understanding properties or behaviors of weathering steel,and reducing the resort to physical testing.展开更多
The process of segmenting point cloud data into several homogeneous areas with points in the same region having the same attributes is known as 3D segmentation.Segmentation is challenging with point cloud data due to...The process of segmenting point cloud data into several homogeneous areas with points in the same region having the same attributes is known as 3D segmentation.Segmentation is challenging with point cloud data due to substantial redundancy,fluctuating sample density and lack of apparent organization.The research area has a wide range of robotics applications,including intelligent vehicles,autonomous mapping and navigation.A number of researchers have introduced various methodologies and algorithms.Deep learning has been successfully used to a spectrum of 2D vision domains as a prevailing A.I.methods.However,due to the specific problems of processing point clouds with deep neural networks,deep learning on point clouds is still in its initial stages.This study examines many strategies that have been presented to 3D instance and semantic segmentation and gives a complete assessment of current developments in deep learning-based 3D segmentation.In these approaches’benefits,draw backs,and design mechanisms are studied and addressed.This study evaluates the impact of various segmentation algorithms on competitiveness on various publicly accessible datasets,as well as the most often used pipelines,their advantages and limits,insightful findings and intriguing future research directions.展开更多
Organoids possess immense potential for unraveling the intricate functions of human tissues and facilitating preclinical disease treatment.Their applications span from high-throughput drug screening to the modeling of...Organoids possess immense potential for unraveling the intricate functions of human tissues and facilitating preclinical disease treatment.Their applications span from high-throughput drug screening to the modeling of complex diseases,with some even achieving clinical translation.Changes in the overall size,shape,boundary,and other morphological features of organoids provide a noninvasive method for assessing organoid drug sensitivity.However,the precise segmentation of organoids in bright-field microscopy images is made difficult by the complexity of the organoid morphology and interference,including overlapping organoids,bubbles,dust particles,and cell fragments.This paper introduces the precision organoid segmentation technique(POST),which is a deep-learning algorithm for segmenting challenging organoids under simple bright-field imaging conditions.Unlike existing methods,POST accurately segments each organoid and eliminates various artifacts encountered during organoid culturing and imaging.Furthermore,it is sensitive to and aligns with measurements of organoid activity in drug sensitivity experiments.POST is expected to be a valuable tool for drug screening using organoids owing to its capability of automatically and rapidly eliminating interfering substances and thereby streamlining the organoid analysis and drug screening process.展开更多
Advanced traffic monitoring systems encounter substantial challenges in vehicle detection and classification due to the limitations of conventional methods,which often demand extensive computational resources and stru...Advanced traffic monitoring systems encounter substantial challenges in vehicle detection and classification due to the limitations of conventional methods,which often demand extensive computational resources and struggle with diverse data acquisition techniques.This research presents a novel approach for vehicle classification and recognition in aerial image sequences,integrating multiple advanced techniques to enhance detection accuracy.The proposed model begins with preprocessing using Multiscale Retinex(MSR)to enhance image quality,followed by Expectation-Maximization(EM)Segmentation for precise foreground object identification.Vehicle detection is performed using the state-of-the-art YOLOv10 framework,while feature extraction incorporates Maximally Stable Extremal Regions(MSER),Dense Scale-Invariant Feature Transform(Dense SIFT),and Zernike Moments Features to capture distinct object characteristics.Feature optimization is further refined through a Hybrid Swarm-based Optimization algorithm,ensuring optimal feature selection for improved classification performance.The final classification is conducted using a Vision Transformer,leveraging its robust learning capabilities for enhanced accuracy.Experimental evaluations on benchmark datasets,including UAVDT and the Unmanned Aerial Vehicle Intruder Dataset(UAVID),demonstrate the superiority of the proposed approach,achieving an accuracy of 94.40%on UAVDT and 93.57%on UAVID.The results highlight the efficacy of the model in significantly enhancing vehicle detection and classification in aerial imagery,outperforming existing methodologies and offering a statistically validated improvement for intelligent traffic monitoring systems compared to existing approaches.展开更多
Accurate segmentation of breast cancer in mammogram images plays a critical role in early diagnosis and treatment planning.As research in this domain continues to expand,various segmentation techniques have been propo...Accurate segmentation of breast cancer in mammogram images plays a critical role in early diagnosis and treatment planning.As research in this domain continues to expand,various segmentation techniques have been proposed across classical image processing,machine learning(ML),deep learning(DL),and hybrid/ensemble models.This study conducts a systematic literature review using the PRISMA methodology,analyzing 57 selected articles to explore how these methods have evolved and been applied.The review highlights the strengths and limitations of each approach,identifies commonly used public datasets,and observes emerging trends in model integration and clinical relevance.By synthesizing current findings,this work provides a structured overview of segmentation strategies and outlines key considerations for developing more adaptable and explainable tools for breast cancer detection.Overall,our synthesis suggests that classical and ML methods are suitable for limited labels and computing resources,while DL models are preferable when pixel-level annotations and resources are available,and hybrid pipelines are most appropriate when fine-grained clinical precision is required.展开更多
This study aimed to enhance the performance of semantic segmentation for autonomous driving by improving the 2DPASS model.Two novel improvements were proposed and implemented in this paper:dynamically adjusting the lo...This study aimed to enhance the performance of semantic segmentation for autonomous driving by improving the 2DPASS model.Two novel improvements were proposed and implemented in this paper:dynamically adjusting the loss function ratio and integrating an attention mechanism(CBAM).First,the loss function weights were adjusted dynamically.The grid search method is used for deciding the best ratio of 7:3.It gives greater emphasis to the cross-entropy loss,which resulted in better segmentation performance.Second,CBAM was applied at different layers of the 2Dencoder.Heatmap analysis revealed that introducing it after the second block of 2D image encoding produced the most effective enhancement of important feature representation.The training epoch was chosen for optimizing the best value by experiments,which improved model convergence and overall accuracy.To evaluate the proposed approach,experiments were conducted based on the SemanticKITTI database.The results showed that the improved model achieved higher segmentation accuracy by 64.31%,improved 11.47% in mIoU compared with the conventional 2DPASS model(baseline:52.84%).It was more effective at detecting small and distant objects and clearly identifying boundaries between different classes.Issues such as noise and variations in data distribution affected its accuracy,indicating the need for further refinement.Overall,the proposed improvements to the 2DPASS model demonstrated the potential to advance semantic segmentation technology and contributed to a more reliable perception of complex,dynamic environments in autonomous vehicles.Accurate segmentation enhances the vehicle’s ability to distinguish different objects,and this improvement directly supports safer navigation,robust decision-making,and efficient path planning,making it highly applicable to real-world deployment of autonomous systems in urban and highway settings.展开更多
Brain tumors require precise segmentation for diagnosis and treatment plans due to their complex morphology and heterogeneous characteristics.While MRI-based automatic brain tumor segmentation technology reduces the b...Brain tumors require precise segmentation for diagnosis and treatment plans due to their complex morphology and heterogeneous characteristics.While MRI-based automatic brain tumor segmentation technology reduces the burden on medical staff and provides quantitative information,existing methodologies and recent models still struggle to accurately capture and classify the fine boundaries and diverse morphologies of tumors.In order to address these challenges and maximize the performance of brain tumor segmentation,this research introduces a novel SwinUNETR-based model by integrating a new decoder block,the Hierarchical Channel-wise Attention Decoder(HCAD),into a powerful SwinUNETR encoder.The HCAD decoder block utilizes hierarchical features and channelspecific attention mechanisms to further fuse information at different scales transmitted from the encoder and preserve spatial details throughout the reconstruction phase.Rigorous evaluations on the recent BraTS GLI datasets demonstrate that the proposed SwinHCAD model achieved superior and improved segmentation accuracy on both the Dice score and HD95 metrics across all tumor subregions(WT,TC,and ET)compared to baseline models.In particular,the rationale and contribution of the model design were clarified through ablation studies to verify the effectiveness of the proposed HCAD decoder block.The results of this study are expected to greatly contribute to enhancing the efficiency of clinical diagnosis and treatment planning by increasing the precision of automated brain tumor segmentation.展开更多
Semantic segmentation of eye images is a complex task with important applications in human–computer interaction,cognitive science,and neuroscience.Achieving real-time,accurate,and robust segmentation algorithms is cr...Semantic segmentation of eye images is a complex task with important applications in human–computer interaction,cognitive science,and neuroscience.Achieving real-time,accurate,and robust segmentation algorithms is crucial for computationally limited portable devices such as augmented reality and virtual reality.With the rapid advancements in deep learning,many network models have been developed specifically for eye image segmentation.Some methods divide the segmentation process into multiple stages to achieve model parameter miniaturization while enhancing output through post processing techniques to improve segmentation accuracy.These approaches significantly increase the inference time.Other networks adopt more complex encoding and decoding modules to achieve end-to-end output,which requires substantial computation.Therefore,balancing the model’s size,accuracy,and computational complexity is essential.To address these challenges,we propose a lightweight asymmetric UNet architecture and a projection loss function.We utilize ResNet-3 layer blocks to enhance feature extraction efficiency in the encoding stage.In the decoding stage,we employ regular convolutions and skip connections to upscale the feature maps from the latent space to the original image size,balancing the model size and segmentation accuracy.In addition,we leverage the geometric features of the eye region and design a projection loss function to further improve the segmentation accuracy without adding any additional inference computational cost.We validate our approach on the OpenEDS2019 dataset for virtual reality and achieve state-of-the-art performance with 95.33%mean intersection over union(mIoU).Our model has only 0.63M parameters and 350 FPS,which are 68%and 200%of the state-of-the-art model RITNet,respectively.展开更多
Segmenting a breast ultrasound image is still challenging due to the presence of speckle noise,dependency on the operator,and the variation of image quality.This paper presents the UltraSegNet architecture that addres...Segmenting a breast ultrasound image is still challenging due to the presence of speckle noise,dependency on the operator,and the variation of image quality.This paper presents the UltraSegNet architecture that addresses these challenges through three key technical innovations:This work adds three things:(1)a changed ResNet-50 backbone with sequential 3×3 convolutions to keep fine anatomical details that are needed for finding lesion boundaries;(2)a computationally efficient regional attention mechanism that works on high-resolution features without using a transformer’s extra memory;and(3)an adaptive feature fusion strategy that changes local and global featuresbasedonhowthe image isbeing used.Extensive evaluation on two distinct datasets demonstrates UltraSegNet’s superior performance:On the BUSI dataset,it obtains a precision of 0.915,a recall of 0.908,and an F1 score of 0.911.In the UDAIT dataset,it achieves robust performance across the board,with a precision of 0.901 and recall of 0.894.Importantly,these improvements are achieved at clinically feasible computation times,taking 235 ms per image on standard GPU hardware.Notably,UltraSegNet does amazingly well on difficult small lesions(less than 10 mm),achieving a detection accuracy of 0.891.This is a huge improvement over traditional methods that have a hard time with small-scale features,as standard models can only achieve 0.63–0.71 accuracy.This improvement in small lesion detection is particularly crucial for early-stage breast cancer identification.Results from this work demonstrate that UltraSegNet can be practically deployable in clinical workflows to improve breast cancer screening accuracy.展开更多
Visual question answering(VQA)is a multimodal task,involving a deep understanding of the image scene and the question’s meaning and capturing the relevant correlations between both modalities to infer the appropriate...Visual question answering(VQA)is a multimodal task,involving a deep understanding of the image scene and the question’s meaning and capturing the relevant correlations between both modalities to infer the appropriate answer.In this paper,we propose a VQA system intended to answer yes/no questions about real-world images,in Arabic.To support a robust VQA system,we work in two directions:(1)Using deep neural networks to semantically represent the given image and question in a fine-grainedmanner,namely ResNet-152 and Gated Recurrent Units(GRU).(2)Studying the role of the utilizedmultimodal bilinear pooling fusion technique in the trade-o.between the model complexity and the overall model performance.Some fusion techniques could significantly increase the model complexity,which seriously limits their applicability for VQA models.So far,there is no evidence of how efficient these multimodal bilinear pooling fusion techniques are for VQA systems dedicated to yes/no questions.Hence,a comparative analysis is conducted between eight bilinear pooling fusion techniques,in terms of their ability to reduce themodel complexity and improve themodel performance in this case of VQA systems.Experiments indicate that these multimodal bilinear pooling fusion techniques have improved the VQA model’s performance,until reaching the best performance of 89.25%.Further,experiments have proven that the number of answers in the developed VQA system is a critical factor that a.ects the effectiveness of these multimodal bilinear pooling techniques in achieving their main objective of reducing the model complexity.The Multimodal Local Perception Bilinear Pooling(MLPB)technique has shown the best balance between the model complexity and its performance,for VQA systems designed to answer yes/no questions.展开更多
BACKGROUND Upper gastrointestinal(UGI)diseases present diagnostic challenges during endoscopy due to visual similarities,indistinct boundaries,and observer variability,which can lead to missed diagnoses and delayed tr...BACKGROUND Upper gastrointestinal(UGI)diseases present diagnostic challenges during endoscopy due to visual similarities,indistinct boundaries,and observer variability,which can lead to missed diagnoses and delayed treatment.Automated segmentation using deep learning(DL)models offers the potential to assist endoscopists,improve diagnostic accuracy,and reduce workload.However,multi-class UGI disease segmentation remains underexplored,with limited annotated datasets and insufficient focus on clinical validation.This study hypothesizes that comparative analysis of different DL architectures can identify models suitable for clinical application,providing actionable insights to reduce diagnostic errors and support clinical decision-making in endoscopic practice.AIM To evaluate 17 state-of-the-art DL models for multi-class UGI disease segmentation,emphasizing clinical translation and real-world applicability.METHODS This study evaluated 17 DL models spanning convolutional neural network(CNN)-,transformer-,and mambabased architectures using a self-collected dataset from two hospitals in Macao and Xiangyang(3313 images,9 classes)and the public EDD2020 dataset(386 images,5 classes).Models were assessed for segmentation performance and performance-efficiency trade-off.Statistical analyses were conducted to examine performance differences across architectures.Generalization capability was measured through a cross-dataset evaluation(training models on the self-collected dataset and testing on the EDD2020 dataset).RESULTS Swin-UMamba achieved the highest segmentation performance across both datasets[intersection over union(IoU):89.06%±0.20%self-collected,77.53%±0.32%EDD2020],followed by SegFormer(IoU:88.94%±0.38%selfcollected,77.20%±0.98%EDD2020)and ConvNeXt+UPerNet(IoU:88.48%±0.09%self-collected,76.90%±0.61%EDD2020).Statistical analyses showed no significant differences between paradigms,though hierarchical architectures with pre-trained encoders consistently outperformed simpler designs.SegFormer achieved the best balance of accuracy and computational efficiency with a performance-efficiency trade-off score of 92.02%,making it suitable for real-time clinical use.Cross-dataset evaluation revealed significant performance drops,with generalization retention rates of 64.78%to 71.52%.Transformer-based models,particularly pyramid vision transformer v2+efficient multi-scale convolutional decoding(IoU:63.35%±1.44%),generalized better than CNN-and mambabased models.CONCLUSION Hierarchical architectures like Swin-UMamba and SegFormer show promise for UGI disease segmentation,reducing missed diagnoses and improving workflows,but robust clinical validation is crucial for real-world deployment.展开更多
Deep learning(DL),derived from the domain of Artificial Neural Networks(ANN),forms one of the most essential components of modern deep learning algorithms.DL segmentation models rely on layer-by-layer convolution-base...Deep learning(DL),derived from the domain of Artificial Neural Networks(ANN),forms one of the most essential components of modern deep learning algorithms.DL segmentation models rely on layer-by-layer convolution-based feature representation,guided by forward and backward propagation.Acritical aspect of this process is the selection of an appropriate activation function(AF)to ensure robustmodel learning.However,existing activation functions often fail to effectively address the vanishing gradient problem or are complicated by the need for manual parameter tuning.Most current research on activation function design focuses on classification tasks using natural image datasets such asMNIST,CIFAR-10,and CIFAR-100.To address this gap,this study proposesMed-ReLU,a novel activation function specifically designed for medical image segmentation.Med-ReLU prevents deep learning models fromsuffering dead neurons or vanishing gradient issues.It is a hybrid activation function that combines the properties of ReLU and Softsign.For positive inputs,Med-ReLU adopts the linear behavior of ReLU to avoid vanishing gradients,while for negative inputs,it exhibits the Softsign’s polynomial convergence,ensuring robust training and avoiding inactive neurons across the training set.The training performance and segmentation accuracy ofMed-ReLU have been thoroughly evaluated,demonstrating stable learning behavior and resistance to overfitting.It consistently outperforms state-of-the-art activation functions inmedical image segmentation tasks.Designed as a parameter-free function,Med-ReLU is simple to implement in complex deep learning architectures,and its effectiveness spans various neural network models and anomaly detection scenarios.展开更多
文摘Modern manufacturing processes have become more reliant on automation because of the accelerated transition from Industry 3.0 to Industry 4.0.Manual inspection of products on assembly lines remains inefficient,prone to errors and lacks consistency,emphasizing the need for a reliable and automated inspection system.Leveraging both object detection and image segmentation approaches,this research proposes a vision-based solution for the detection of various kinds of tools in the toolkit using deep learning(DL)models.Two Intel RealSense D455f depth cameras were arranged in a top down configuration to capture both RGB and depth images of the toolkits.After applying multiple constraints and enhancing them through preprocessing and augmentation,a dataset consisting of 3300 annotated RGB-D photos was generated.Several DL models were selected through a comprehensive assessment of mean Average Precision(mAP),precision-recall equilibrium,inference latency(target≥30 FPS),and computational burden,resulting in a preference for YOLO and Region-based Convolutional Neural Networks(R-CNN)variants over ViT-based models due to the latter’s increased latency and resource requirements.YOLOV5,YOLOV8,YOLOV11,Faster R-CNN,and Mask R-CNN were trained on the annotated dataset and evaluated using key performance metrics(Recall,Accuracy,F1-score,and Precision).YOLOV11 demonstrated balanced excellence with 93.0%precision,89.9%recall,and a 90.6%F1-score in object detection,as well as 96.9%precision,95.3%recall,and a 96.5%F1-score in instance segmentation with an average inference time of 25 ms per frame(≈40 FPS),demonstrating real-time performance.Leveraging these results,a YOLOV11-based windows application was successfully deployed in a real-time assembly line environment,where it accurately processed live video streams to detect and segment tools within toolkits,demonstrating its practical effectiveness in industrial automation.The application is capable of precisely measuring socket dimensions by utilising edge detection techniques on YOLOv11 segmentation masks,in addition to detection and segmentation.This makes it possible to do specification-level quality control right on the assembly line,which improves the ability to examine things in real time.The implementation is a big step forward for intelligent manufacturing in the Industry 4.0 paradigm.It provides a scalable,efficient,and accurate way to do automated inspection and dimensional verification activities.
基金financially supported by the Open Project Program of Wuhan National Laboratory for Optoelectronics(No.2022WNLOKF009)the National Natural Science Foundation of China(No.62475216)+2 种基金the Key Research and Development Program of Shaanxi(No.2024GH-ZDXM-37)the Fujian Provincial Natural Science Foundation of China(No.2024J01060)the Startup Program of XMU,and the Fundamental Research Funds for the Central Universities.
文摘Microscopy imaging is fundamental in analyzing bacterial morphology and dynamics,offering critical insights into bacterial physiology and pathogenicity.Image segmentation techniques enable quantitative analysis of bacterial structures,facilitating precise measurement of morphological variations and population behaviors at single-cell resolution.This paper reviews advancements in bacterial image segmentation,emphasizing the shift from traditional thresholding and watershed methods to deep learning-driven approaches.Convolutional neural networks(CNNs),U-Net architectures,and three-dimensional(3D)frameworks excel at segmenting dense biofilms and resolving antibiotic-induced morphological changes.These methods combine automated feature extraction with physics-informed postprocessing.Despite progress,challenges persist in computational efficiency,cross-species generalizability,and integration with multimodal experimental workflows.Future progress will depend on improving model robustness across species and imaging modalities,integrating multimodal data for phenotype-function mapping,and developing standard pipelines that link computational tools with clinical diagnostics.These innovations will expand microbial phenotyping beyond structural analysis,enabling deeper insights into bacterial physiology and ecological interactions.
文摘This systematic review aims to comprehensively examine and compare deep learning methods for brain tumor segmentation and classification using MRI and other imaging modalities,focusing on recent trends from 2022 to 2025.The primary objective is to evaluate methodological advancements,model performance,dataset usage,and existing challenges in developing clinically robust AI systems.We included peer-reviewed journal articles and highimpact conference papers published between 2022 and 2025,written in English,that proposed or evaluated deep learning methods for brain tumor segmentation and/or classification.Excluded were non-open-access publications,books,and non-English articles.A structured search was conducted across Scopus,Google Scholar,Wiley,and Taylor&Francis,with the last search performed in August 2025.Risk of bias was not formally quantified but considered during full-text screening based on dataset diversity,validation methods,and availability of performance metrics.We used narrative synthesis and tabular benchmarking to compare performance metrics(e.g.,accuracy,Dice score)across model types(CNN,Transformer,Hybrid),imaging modalities,and datasets.A total of 49 studies were included(43 journal articles and 6 conference papers).These studies spanned over 9 public datasets(e.g.,BraTS,Figshare,REMBRANDT,MOLAB)and utilized a range of imaging modalities,predominantly MRI.Hybrid models,especially ResViT and UNetFormer,consistently achieved high performance,with classification accuracy exceeding 98%and segmentation Dice scores above 0.90 across multiple studies.Transformers and hybrid architectures showed increasing adoption post2023.Many studies lacked external validation and were evaluated only on a few benchmark datasets,raising concerns about generalizability and dataset bias.Few studies addressed clinical interpretability or uncertainty quantification.Despite promising results,particularly for hybrid deep learning models,widespread clinical adoption remains limited due to lack of validation,interpretability concerns,and real-world deployment barriers.
基金supported by the National Natural Science Foundation of China[grant number 62376217]the Young Elite Scientists Sponsorship Program by CAST[grant number 2023QNRC001]the Joint Research Project for Meteorological Capacity Improvement[grant number 24NLTSZ003]。
文摘Deep learning-based methods have become alternatives to traditional numerical weather prediction systems,offering faster computation and the ability to utilize large historical datasets.However,the application of deep learning to medium-range regional weather forecasting with limited data remains a significant challenge.In this work,three key solutions are proposed:(1)motivated by the need to improve model performance in data-scarce regional forecasting scenarios,the authors innovatively apply semantic segmentation models,to better capture spatiotemporal features and improve prediction accuracy;(2)recognizing the challenge of overfitting and the inability of traditional noise-based data augmentation methods to effectively enhance model robustness,a novel learnable Gaussian noise mechanism is introduced that allows the model to adaptively optimize perturbations for different locations,ensuring more effective learning;and(3)to address the issue of error accumulation in autoregressive prediction,as well as the challenge of learning difficulty and the lack of intermediate data utilization in one-shot prediction,the authors propose a cascade prediction approach that effectively resolves these problems while significantly improving model forecasting performance.The method achieves a competitive result in The East China Regional AI Medium Range Weather Forecasting Competition.Ablation experiments further validate the effectiveness of each component,highlighting their contributions to enhancing prediction performance.
基金supported by the National Research Foundation of Korea(NRF)grant funded by theKorea government(MSIT)(No.RS-2024-00405278)partially supported by the Jeju Industry-University Convergence District Project for Promoting Industry-Campus Cooperationfunded by the Ministry of Trade,Industry and Energy(MOTIE,Korea)[Project Name:Jeju Industry-University Convergence District Project for Promoting Industry-Campus Cooperation/Project Number:P0029950].
文摘Recent advances in deep learning have significantly improved flood detection and segmentation from aerial and satellite imagery.However,conventional convolutional neural networks(CNNs)often struggle in complex flood scenarios involving reflections,occlusions,or indistinct boundaries due to limited contextual modeling.To address these challenges,we propose a hybrid flood segmentation framework that integrates a Vision Transformer(ViT)encoder with a U-Net decoder,enhanced by a novel Flood-Aware Refinement Block(FARB).The FARB module improves boundary delineation and suppresses noise by combining residual smoothing with spatial-channel attention mechanisms.We evaluate our model on a UAV-acquired flood imagery dataset,demonstrating that the proposed ViTUNet+FARB architecture outperforms existing CNN and Transformer-based models in terms of accuracy and mean Intersection over Union(mIoU).Detailed ablation studies further validate the contribution of each component,confirming that the FARB design significantly enhances segmentation quality.To its better performance and computational efficiency,the proposed framework is well-suited for flood monitoring and disaster response applications,particularly in resource-constrained environments.
基金supported in part by the National Natural Science Foundation of China under Grant 62071405the National Natural Science Foundation of China under Grant 12175189.
文摘The accurate segmentation of deep gray matter nuclei is critical for neuropathological research,disease diagnosis and treatment.Existing methods employ the supervised learning training approach,which requires large labeled datasets.It is challenging and time-consuming to obtain such datasets for medical image analysis.In addition,these methods based on convolutional neural networks(CNNs)only achieve suboptimal performance due to the locality of convolutional operations.Vision Transformers(ViTs)efficiently model long-range dependencies and thus have the potentiality to outperform these methods in segmentation tasks.To address these issues,we propose a novel hybrid network based on self-supervised pre-training for deep gray matter nuclei segmentation.Specifically,we present a CNN-Transformer hybrid network(CTNet),whose encoder consists of 3D CNN and ViT to learn local spatial-detailed features and global semantic information.A self-supervised learning(SSL)approach that integrates rotation prediction and masked feature reconstruction is proposed to pre-train the CTNet,enabling the model to learn valuable visual representations from unlabeled data.We evaluate the effectiveness of our method on 3T and 7T human brain MRI datasets.The results demonstrate that our CTNet achieves better performance than other comparison models and our pre-training strategy outperforms other advanced self-supervised methods.When the training set has only one sample,our pre-trained CTNet enhances segmentation performance,showing an 8.4%improvement in Dice similarity coefficient(DSC)compared to the randomly initialized CTNet.
基金funded by the Ongoing Research Funding Program-Research Chairs(ORF-RC-2025-2400),King Saud University,Riyadh,Saudi Arabia。
文摘Recent studies indicate that millions of individuals suffer from renal diseases,with renal carcinoma,a type of kidney cancer,emerging as both a chronic illness and a significant cause of mortality.Magnetic Resonance Imaging(MRI)and Computed Tomography(CT)have become essential tools for diagnosing and assessing kidney disorders.However,accurate analysis of thesemedical images is critical for detecting and evaluating tumor severity.This study introduces an integrated hybrid framework that combines three complementary deep learning models for kidney tumor segmentation from MRI images.The proposed framework fuses a customized U-Net and Mask R-CNN using a weighted scheme to achieve semantic and instance-level segmentation.The fused outputs are further refined through edge detection using Stochastic FeatureMapping Neural Networks(SFMNN),while volumetric consistency is ensured through Improved Mini-Batch K-Means(IMBKM)clustering integrated with an Encoder-Decoder Convolutional Neural Network(EDCNN).The outputs of these three stages are combined through a weighted fusion mechanism,with optimal weights determined empirically.Experiments on MRI scans from the TCGA-KIRC dataset demonstrate that the proposed hybrid framework significantly outperforms standalone models,achieving a Dice Score of 92.5%,an IoU of 87.8%,a Precision of 93.1%,a Recall of 90.8%,and a Hausdorff Distance of 2.8 mm.These findings validate that the weighted integration of complementary architectures effectively overcomes key limitations in kidney tumor segmentation,leading to improved diagnostic accuracy and robustness in medical image analysis.
基金supported by the National Natural Science Foundation of China(Nos.61373121 and 61328205)Program for Sichuan Provincial Science Fund for Distinguished Young Scholars(No.13QNJJ0149)+1 种基金the Fundamental Research Funds for the Central UniversitiesChina Scholarship Council(No.201507000032)
文摘The deep learning technology has shown impressive performance in various vision tasks such as image classification, object detection and semantic segmentation. In particular, recent advances of deep learning techniques bring encouraging performance to fine-grained image classification which aims to distinguish subordinate-level categories, such as bird species or dog breeds. This task is extremely challenging due to high intra-class and low inter-class variance. In this paper, we review four types of deep learning based fine-grained image classification approaches, including the general convolutional neural networks （CNNs）, part detection based, ensemble of networks based and visual attention based fine-grained image classification approaches. Besides, the deep learning based semantic segmentation approaches are also covered in this paper. The region proposal based and fully convolutional networks based approaches for semantic segmentation are introduced respectively.
基金supported by the National Key Research and Development Program of China(No.2017YFB0702303).
文摘Microstructural classification is typically done manually by human experts,which gives rise to uncertainties due to subjectivity and reduces the overall efficiency.A high-throughput characterization is proposed based on deep learning,rapid acquisition technology,and mathematical statistics for the recognition,segmentation,and quantification of microstructure in weathering steel.The segmentation results showed that this method was accurate and efficient,and the segmentation of inclusions and pearlite phase achieved accuracy of 89.95%and 90.86%,respectively.The time required for batch processing by MIPAR software involving thresholding segmentation,morphological processing,and small area deletion was 1.05 s for a single image.By comparison,our system required only 0.102 s,which is ten times faster than the commercial software.The quantification results were extracted from large volumes of sequential image data(150 mm^(2),62,216 images,1024×1024 pixels),which ensure comprehensive statistics.Microstructure information,such as three-dimensional density distribution and the frequency of the minimum spatial distance of inclusions on the sample surface of 150 mm^(2),were quantified by extracting the coordinates and sizes of individual features.A refined characterization method for two-dimensional structures and spatial information that is unattainable when performing manually or with software is provided.That will be useful for understanding properties or behaviors of weathering steel,and reducing the resort to physical testing.
基金This research was supported by the BB21 plus funded by Busan Metropolitan City and Busan Institute for Talent and Lifelong Education(BIT)and a grant from Tongmyong University Innovated University Research Park(I-URP)funded by Busan Metropolitan City,Republic of Korea.
文摘The process of segmenting point cloud data into several homogeneous areas with points in the same region having the same attributes is known as 3D segmentation.Segmentation is challenging with point cloud data due to substantial redundancy,fluctuating sample density and lack of apparent organization.The research area has a wide range of robotics applications,including intelligent vehicles,autonomous mapping and navigation.A number of researchers have introduced various methodologies and algorithms.Deep learning has been successfully used to a spectrum of 2D vision domains as a prevailing A.I.methods.However,due to the specific problems of processing point clouds with deep neural networks,deep learning on point clouds is still in its initial stages.This study examines many strategies that have been presented to 3D instance and semantic segmentation and gives a complete assessment of current developments in deep learning-based 3D segmentation.In these approaches’benefits,draw backs,and design mechanisms are studied and addressed.This study evaluates the impact of various segmentation algorithms on competitiveness on various publicly accessible datasets,as well as the most often used pipelines,their advantages and limits,insightful findings and intriguing future research directions.
基金supported by the National Key R&D Program of China(No.2022YFC2504403)the National Natural Science Foundation of China(No.62172202)+1 种基金the Experiment Project of China Manned Space Program(No.HYZHXM01019)the Fundamental Research Funds for the Central Universities from Southeast University(No.3207032101C3)。
文摘Organoids possess immense potential for unraveling the intricate functions of human tissues and facilitating preclinical disease treatment.Their applications span from high-throughput drug screening to the modeling of complex diseases,with some even achieving clinical translation.Changes in the overall size,shape,boundary,and other morphological features of organoids provide a noninvasive method for assessing organoid drug sensitivity.However,the precise segmentation of organoids in bright-field microscopy images is made difficult by the complexity of the organoid morphology and interference,including overlapping organoids,bubbles,dust particles,and cell fragments.This paper introduces the precision organoid segmentation technique(POST),which is a deep-learning algorithm for segmenting challenging organoids under simple bright-field imaging conditions.Unlike existing methods,POST accurately segments each organoid and eliminates various artifacts encountered during organoid culturing and imaging.Furthermore,it is sensitive to and aligns with measurements of organoid activity in drug sensitivity experiments.POST is expected to be a valuable tool for drug screening using organoids owing to its capability of automatically and rapidly eliminating interfering substances and thereby streamlining the organoid analysis and drug screening process.
文摘Advanced traffic monitoring systems encounter substantial challenges in vehicle detection and classification due to the limitations of conventional methods,which often demand extensive computational resources and struggle with diverse data acquisition techniques.This research presents a novel approach for vehicle classification and recognition in aerial image sequences,integrating multiple advanced techniques to enhance detection accuracy.The proposed model begins with preprocessing using Multiscale Retinex(MSR)to enhance image quality,followed by Expectation-Maximization(EM)Segmentation for precise foreground object identification.Vehicle detection is performed using the state-of-the-art YOLOv10 framework,while feature extraction incorporates Maximally Stable Extremal Regions(MSER),Dense Scale-Invariant Feature Transform(Dense SIFT),and Zernike Moments Features to capture distinct object characteristics.Feature optimization is further refined through a Hybrid Swarm-based Optimization algorithm,ensuring optimal feature selection for improved classification performance.The final classification is conducted using a Vision Transformer,leveraging its robust learning capabilities for enhanced accuracy.Experimental evaluations on benchmark datasets,including UAVDT and the Unmanned Aerial Vehicle Intruder Dataset(UAVID),demonstrate the superiority of the proposed approach,achieving an accuracy of 94.40%on UAVDT and 93.57%on UAVID.The results highlight the efficacy of the model in significantly enhancing vehicle detection and classification in aerial imagery,outperforming existing methodologies and offering a statistically validated improvement for intelligent traffic monitoring systems compared to existing approaches.
基金funded by BK21 FOUR(Fostering Outstanding Universities for Research)(No.:5199990914048).
文摘Accurate segmentation of breast cancer in mammogram images plays a critical role in early diagnosis and treatment planning.As research in this domain continues to expand,various segmentation techniques have been proposed across classical image processing,machine learning(ML),deep learning(DL),and hybrid/ensemble models.This study conducts a systematic literature review using the PRISMA methodology,analyzing 57 selected articles to explore how these methods have evolved and been applied.The review highlights the strengths and limitations of each approach,identifies commonly used public datasets,and observes emerging trends in model integration and clinical relevance.By synthesizing current findings,this work provides a structured overview of segmentation strategies and outlines key considerations for developing more adaptable and explainable tools for breast cancer detection.Overall,our synthesis suggests that classical and ML methods are suitable for limited labels and computing resources,while DL models are preferable when pixel-level annotations and resources are available,and hybrid pipelines are most appropriate when fine-grained clinical precision is required.
文摘This study aimed to enhance the performance of semantic segmentation for autonomous driving by improving the 2DPASS model.Two novel improvements were proposed and implemented in this paper:dynamically adjusting the loss function ratio and integrating an attention mechanism(CBAM).First,the loss function weights were adjusted dynamically.The grid search method is used for deciding the best ratio of 7:3.It gives greater emphasis to the cross-entropy loss,which resulted in better segmentation performance.Second,CBAM was applied at different layers of the 2Dencoder.Heatmap analysis revealed that introducing it after the second block of 2D image encoding produced the most effective enhancement of important feature representation.The training epoch was chosen for optimizing the best value by experiments,which improved model convergence and overall accuracy.To evaluate the proposed approach,experiments were conducted based on the SemanticKITTI database.The results showed that the improved model achieved higher segmentation accuracy by 64.31%,improved 11.47% in mIoU compared with the conventional 2DPASS model(baseline:52.84%).It was more effective at detecting small and distant objects and clearly identifying boundaries between different classes.Issues such as noise and variations in data distribution affected its accuracy,indicating the need for further refinement.Overall,the proposed improvements to the 2DPASS model demonstrated the potential to advance semantic segmentation technology and contributed to a more reliable perception of complex,dynamic environments in autonomous vehicles.Accurate segmentation enhances the vehicle’s ability to distinguish different objects,and this improvement directly supports safer navigation,robust decision-making,and efficient path planning,making it highly applicable to real-world deployment of autonomous systems in urban and highway settings.
基金supported by Institute of Information&Communications Technology Planning&Evaluation(IITP)under the Metaverse Support Program to Nurture the Best Talents(IITP-2024-RS-2023-00254529)grant funded by the Korea government(MSIT).
文摘Brain tumors require precise segmentation for diagnosis and treatment plans due to their complex morphology and heterogeneous characteristics.While MRI-based automatic brain tumor segmentation technology reduces the burden on medical staff and provides quantitative information,existing methodologies and recent models still struggle to accurately capture and classify the fine boundaries and diverse morphologies of tumors.In order to address these challenges and maximize the performance of brain tumor segmentation,this research introduces a novel SwinUNETR-based model by integrating a new decoder block,the Hierarchical Channel-wise Attention Decoder(HCAD),into a powerful SwinUNETR encoder.The HCAD decoder block utilizes hierarchical features and channelspecific attention mechanisms to further fuse information at different scales transmitted from the encoder and preserve spatial details throughout the reconstruction phase.Rigorous evaluations on the recent BraTS GLI datasets demonstrate that the proposed SwinHCAD model achieved superior and improved segmentation accuracy on both the Dice score and HD95 metrics across all tumor subregions(WT,TC,and ET)compared to baseline models.In particular,the rationale and contribution of the model design were clarified through ablation studies to verify the effectiveness of the proposed HCAD decoder block.The results of this study are expected to greatly contribute to enhancing the efficiency of clinical diagnosis and treatment planning by increasing the precision of automated brain tumor segmentation.
基金supported by the HFIPS Director’s Foundation(YZJJ202207-TS),the National Natural Science Foundation of China(82371931)the Natural Science Foundation of Anhui Province(2008085MC69)+3 种基金the Natural Science Foundation of Hefei City(2021033)the General Scientific Research Project of Anhui Provincial Health Commission(AHWJ2021b150)the Collaborative Innovation Program of Hefei Science Center,CAS(2021HSC-CIP013)the Anhui Province Key Research and Development Project(202204295107020004).
文摘Semantic segmentation of eye images is a complex task with important applications in human–computer interaction,cognitive science,and neuroscience.Achieving real-time,accurate,and robust segmentation algorithms is crucial for computationally limited portable devices such as augmented reality and virtual reality.With the rapid advancements in deep learning,many network models have been developed specifically for eye image segmentation.Some methods divide the segmentation process into multiple stages to achieve model parameter miniaturization while enhancing output through post processing techniques to improve segmentation accuracy.These approaches significantly increase the inference time.Other networks adopt more complex encoding and decoding modules to achieve end-to-end output,which requires substantial computation.Therefore,balancing the model’s size,accuracy,and computational complexity is essential.To address these challenges,we propose a lightweight asymmetric UNet architecture and a projection loss function.We utilize ResNet-3 layer blocks to enhance feature extraction efficiency in the encoding stage.In the decoding stage,we employ regular convolutions and skip connections to upscale the feature maps from the latent space to the original image size,balancing the model size and segmentation accuracy.In addition,we leverage the geometric features of the eye region and design a projection loss function to further improve the segmentation accuracy without adding any additional inference computational cost.We validate our approach on the OpenEDS2019 dataset for virtual reality and achieve state-of-the-art performance with 95.33%mean intersection over union(mIoU).Our model has only 0.63M parameters and 350 FPS,which are 68%and 200%of the state-of-the-art model RITNet,respectively.
基金funded by Princess Nourah bint Abdulrahman University Researchers Supporting Project number(PNURSP2025R435),Princess Nourah bint Abdulrahman University,Riyadh,Saudi Arabia.
文摘Segmenting a breast ultrasound image is still challenging due to the presence of speckle noise,dependency on the operator,and the variation of image quality.This paper presents the UltraSegNet architecture that addresses these challenges through three key technical innovations:This work adds three things:(1)a changed ResNet-50 backbone with sequential 3×3 convolutions to keep fine anatomical details that are needed for finding lesion boundaries;(2)a computationally efficient regional attention mechanism that works on high-resolution features without using a transformer’s extra memory;and(3)an adaptive feature fusion strategy that changes local and global featuresbasedonhowthe image isbeing used.Extensive evaluation on two distinct datasets demonstrates UltraSegNet’s superior performance:On the BUSI dataset,it obtains a precision of 0.915,a recall of 0.908,and an F1 score of 0.911.In the UDAIT dataset,it achieves robust performance across the board,with a precision of 0.901 and recall of 0.894.Importantly,these improvements are achieved at clinically feasible computation times,taking 235 ms per image on standard GPU hardware.Notably,UltraSegNet does amazingly well on difficult small lesions(less than 10 mm),achieving a detection accuracy of 0.891.This is a huge improvement over traditional methods that have a hard time with small-scale features,as standard models can only achieve 0.63–0.71 accuracy.This improvement in small lesion detection is particularly crucial for early-stage breast cancer identification.Results from this work demonstrate that UltraSegNet can be practically deployable in clinical workflows to improve breast cancer screening accuracy.
文摘Visual question answering(VQA)is a multimodal task,involving a deep understanding of the image scene and the question’s meaning and capturing the relevant correlations between both modalities to infer the appropriate answer.In this paper,we propose a VQA system intended to answer yes/no questions about real-world images,in Arabic.To support a robust VQA system,we work in two directions:(1)Using deep neural networks to semantically represent the given image and question in a fine-grainedmanner,namely ResNet-152 and Gated Recurrent Units(GRU).(2)Studying the role of the utilizedmultimodal bilinear pooling fusion technique in the trade-o.between the model complexity and the overall model performance.Some fusion techniques could significantly increase the model complexity,which seriously limits their applicability for VQA models.So far,there is no evidence of how efficient these multimodal bilinear pooling fusion techniques are for VQA systems dedicated to yes/no questions.Hence,a comparative analysis is conducted between eight bilinear pooling fusion techniques,in terms of their ability to reduce themodel complexity and improve themodel performance in this case of VQA systems.Experiments indicate that these multimodal bilinear pooling fusion techniques have improved the VQA model’s performance,until reaching the best performance of 89.25%.Further,experiments have proven that the number of answers in the developed VQA system is a critical factor that a.ects the effectiveness of these multimodal bilinear pooling techniques in achieving their main objective of reducing the model complexity.The Multimodal Local Perception Bilinear Pooling(MLPB)technique has shown the best balance between the model complexity and its performance,for VQA systems designed to answer yes/no questions.
基金Supported by the Guangdong Basic and Applied Basic Research Foundation,No.2021B1515130003the Key Research and Development Plan of Hubei Province,No.2022BCE034the Natural Science Foundation of Hubei Province,No.2024AFB1054.
文摘BACKGROUND Upper gastrointestinal(UGI)diseases present diagnostic challenges during endoscopy due to visual similarities,indistinct boundaries,and observer variability,which can lead to missed diagnoses and delayed treatment.Automated segmentation using deep learning(DL)models offers the potential to assist endoscopists,improve diagnostic accuracy,and reduce workload.However,multi-class UGI disease segmentation remains underexplored,with limited annotated datasets and insufficient focus on clinical validation.This study hypothesizes that comparative analysis of different DL architectures can identify models suitable for clinical application,providing actionable insights to reduce diagnostic errors and support clinical decision-making in endoscopic practice.AIM To evaluate 17 state-of-the-art DL models for multi-class UGI disease segmentation,emphasizing clinical translation and real-world applicability.METHODS This study evaluated 17 DL models spanning convolutional neural network(CNN)-,transformer-,and mambabased architectures using a self-collected dataset from two hospitals in Macao and Xiangyang(3313 images,9 classes)and the public EDD2020 dataset(386 images,5 classes).Models were assessed for segmentation performance and performance-efficiency trade-off.Statistical analyses were conducted to examine performance differences across architectures.Generalization capability was measured through a cross-dataset evaluation(training models on the self-collected dataset and testing on the EDD2020 dataset).RESULTS Swin-UMamba achieved the highest segmentation performance across both datasets[intersection over union(IoU):89.06%±0.20%self-collected,77.53%±0.32%EDD2020],followed by SegFormer(IoU:88.94%±0.38%selfcollected,77.20%±0.98%EDD2020)and ConvNeXt+UPerNet(IoU:88.48%±0.09%self-collected,76.90%±0.61%EDD2020).Statistical analyses showed no significant differences between paradigms,though hierarchical architectures with pre-trained encoders consistently outperformed simpler designs.SegFormer achieved the best balance of accuracy and computational efficiency with a performance-efficiency trade-off score of 92.02%,making it suitable for real-time clinical use.Cross-dataset evaluation revealed significant performance drops,with generalization retention rates of 64.78%to 71.52%.Transformer-based models,particularly pyramid vision transformer v2+efficient multi-scale convolutional decoding(IoU:63.35%±1.44%),generalized better than CNN-and mambabased models.CONCLUSION Hierarchical architectures like Swin-UMamba and SegFormer show promise for UGI disease segmentation,reducing missed diagnoses and improving workflows,but robust clinical validation is crucial for real-world deployment.
基金The researchers would like to thank the Deanship of Graduate Studies and Scientific Research at Qassim University for financial support(QU-APC-2025).
文摘Deep learning(DL),derived from the domain of Artificial Neural Networks(ANN),forms one of the most essential components of modern deep learning algorithms.DL segmentation models rely on layer-by-layer convolution-based feature representation,guided by forward and backward propagation.Acritical aspect of this process is the selection of an appropriate activation function(AF)to ensure robustmodel learning.However,existing activation functions often fail to effectively address the vanishing gradient problem or are complicated by the need for manual parameter tuning.Most current research on activation function design focuses on classification tasks using natural image datasets such asMNIST,CIFAR-10,and CIFAR-100.To address this gap,this study proposesMed-ReLU,a novel activation function specifically designed for medical image segmentation.Med-ReLU prevents deep learning models fromsuffering dead neurons or vanishing gradient issues.It is a hybrid activation function that combines the properties of ReLU and Softsign.For positive inputs,Med-ReLU adopts the linear behavior of ReLU to avoid vanishing gradients,while for negative inputs,it exhibits the Softsign’s polynomial convergence,ensuring robust training and avoiding inactive neurons across the training set.The training performance and segmentation accuracy ofMed-ReLU have been thoroughly evaluated,demonstrating stable learning behavior and resistance to overfitting.It consistently outperforms state-of-the-art activation functions inmedical image segmentation tasks.Designed as a parameter-free function,Med-ReLU is simple to implement in complex deep learning architectures,and its effectiveness spans various neural network models and anomaly detection scenarios.
