Background:The reputation system has been designed as an effective mechanism to reduce risks associated with online shopping for customers.However,it is vulnerable to rating fraud.Some raters may inject unfairly high ...Background:The reputation system has been designed as an effective mechanism to reduce risks associated with online shopping for customers.However,it is vulnerable to rating fraud.Some raters may inject unfairly high or low ratings to the system so as to promote their own products or demote their competitors.Method:This study explores the rating fraud by differentiating the subjective fraud from objective fraud.Then it discusses the effectiveness of blockchain technology in objective fraud and its limitation in subjective fraud,especially the rating fraud.Lastly,it systematically analyzes the robustness of blockchain-based reputation systems in each type of rating fraud.Results:The detection of fraudulent raters is not easy since they can behave strategically to camouflage themselves.We explore the potential strengths and limitations of blockchain-based reputation systems under two attack goals:ballot-stuffing and bad-mouthing,and various attack models including constant attack,camouflage attack,whitewashing attack and sybil attack.Blockchain-based reputation systems are more robust against bad-mouthing than ballot-stuffing fraud.Conclusions:Blockchain technology provides new opportunities for redesigning the reputation system.Blockchain systems are very effective in preventing objective information fraud,such as loan application fraud,where fraudulent information is fact-based.However,their effectiveness is limited in subjective information fraud,such as rating fraud,where the ground-truth is not easily validated.Blockchain systems are effective in preventing bad mouthing and whitewashing attack,but they are limited in detecting ballot-stuffing under sybil attack,constant attacks and camouflage attack.展开更多
With the complexity of integrated circuits is continually increasing, a local defect in circuits may cause multiple faults. The behavior of a digital circuit with a multiple fault may significantly differ from that of...With the complexity of integrated circuits is continually increasing, a local defect in circuits may cause multiple faults. The behavior of a digital circuit with a multiple fault may significantly differ from that of a single fault. A new method for the detection of multiple faults in digital circuits is presented in this paper, the method is based on binary decision diagram （BDD）. First of all, the BDDs for the normal circuit and faulty circuit are built respectively. Secondly, a test BDD is obtained by the XOR operation of the BDDs corresponds to normal circuit and faulty circuit. In the test BDD, each input assignment that leads to the leaf node labeled 1 is a test vector of multiple faults. Therefore, the test set of multiple faults is generated by searching for the type of input assignments in the test BDD. Experimental results on some digital circuits show the feasibility of the approach presented in this paper.展开更多
This study designs a microstrip patch antenna with an inverted T-type notch in the partial ground to detect tumorcells inside the human breast.The size of the current antenna is small enough(18mm×21mm×1.6mm)...This study designs a microstrip patch antenna with an inverted T-type notch in the partial ground to detect tumorcells inside the human breast.The size of the current antenna is small enough(18mm×21mm×1.6mm)todistribute around the breast phantom.The operating frequency has been observed from6–14GHzwith a minimumreturn loss of−61.18 dB and themaximumgain of current proposed antenna is 5.8 dBiwhich is flexiblewith respectto the size of antenna.After the distribution of eight antennas around the breast phantom,the return loss curveswere observed in the presence and absence of tumor cells inside the breast phantom,and these observations showa sharp difference between the presence and absence of tumor cells.The simulated results show that this proposedantenna is suitable for early detection of cancerous cells inside the breast.展开更多
Surface-enhanced Raman scattering(SERS)is a powerful technology for obtaining vibrational information from molecules that present in different chemical or biological environments.This paper presents a 3D SERS substrat...Surface-enhanced Raman scattering(SERS)is a powerful technology for obtaining vibrational information from molecules that present in different chemical or biological environments.This paper presents a 3D SERS substrate based on nanocone forests.The substrates are prepared by using plasma treatment technique,which is a simple,fast and high-throughput approach.The SERS substrate based on nanocone forests exhibits high sensitivity.In the experiment,miRNA with a concentration as low as 10-10 M can be achieved.Meanwhile,the proposed SERS substrate shows a high uniformity over a large area.These experimental results demonstrate great potential of the 3D SERS substrate in wide applications.展开更多
To solve the problem that the conventional detections in DS-CDMA suffer from high complexity and poor robustness for the time-hopping pulse signals, the received pulse signals were remodeled, and a mulfipath-free dete...To solve the problem that the conventional detections in DS-CDMA suffer from high complexity and poor robustness for the time-hopping pulse signals, the received pulse signals were remodeled, and a mulfipath-free detection scheme, which provides a simple approach to select samples of received signals, was introduced. By this scheme, the subsequent multiuser detection （MUD） would get rid of the mis- match due to the correlative multipath signal in IR-UWB. In addition, a computationally efficient recur-sive least squares （RLS） type algorithm based on least mean fourth （LMF） criterion is derived to suppress multi-access interference. The proposed multiuser detection algorithm performs well at low complexity, even in dense muhipath environment.展开更多
Ferroptosis has exhibited great potential in therapies and intracellular reducing agents of sulfur species(RSSs) in the thiol-dependent redox systems are crucial in ferroptosis.This makes the simultaneous detection of...Ferroptosis has exhibited great potential in therapies and intracellular reducing agents of sulfur species(RSSs) in the thiol-dependent redox systems are crucial in ferroptosis.This makes the simultaneous detection of multiple RSSs significant for evaluating ferroptosis therapy.However,the traditional techniques,including fluorescent(FL) imaging and electrospray ionization-based mass spectrometry(MS) detection,cannot achieve the discrimination of different RSSs.Herein,simultaneous MS detection of multiple RSSs,including cysteine(Cys),homocysteine(Hcy),glutathione(GSH) and hydrogen sulfide(H_(2)S),was obtained upon enhancing ionization efficiency by a fluorescent probe(NBD-O-1).Based on the interaction between NBD-O-1 and RSSs,the complex of RSSs with a fragment of NBD-O-1 can be generated,which can be easily ionized for MS detection in the negative mode.Therefore,the intracellular RSSs can be well detected upon the incubation of He La cells with the probe of NBD-O-1,exhibiting the total RSS levels by the FL imaging and further providing expression of each RSS by enhanced MS detection.Furthermore,the RSSs during ferroptosis in He La cells have been evaluated using the present strategy,demonstrating the potential for ferroptosis examinations.This work has made an unconventional application of a fluorescent probe to enhance the detection of multiple RSSs by MS,providing significant molecular information for addressing the ferroptosis mechanism.展开更多
Visible and infrared(RGB-IR)fusion object detection plays an important role in security,disaster relief,etc.In recent years,deep-learning-based RGB-IR fusion detection methods have been developing rapidly,but still st...Visible and infrared(RGB-IR)fusion object detection plays an important role in security,disaster relief,etc.In recent years,deep-learning-based RGB-IR fusion detection methods have been developing rapidly,but still struggle to deal with the complex and changing scenarios captured by drones,mainly due to two reasons:(A)RGB-IR fusion detectors are susceptible to inferior inputs that degrade performance and stability.(B)RGB-IR fusion detectors are susceptible to redundant features that reduce accuracy and efficiency.In this paper,an innovative RGB-IR fusion detection framework based on global-local feature optimization,named GLFDet,is proposed to improve the detection performance and efficiency of drone-captured objects.The key components of GLFDet include a Global Feature Optimization(GFO)module,a Local Feature Optimization(LFO)module and a Channel Separation Fusion(CSF)module.Specifically,GFO calculates the information content of the input image from the frequency domain and optimizes the features holistically.Then,LFO dynamically selects high-value features and filters out low-value features before fusion,which significantly improves the efficiency of fusion.Finally,CSF fuses the RGB and IR features across the corresponding channels,which avoids the rearrangement of the channel relationships and enhances the model stability.Extensive experimental results show that the proposed method achieves the best performance on three popular RGB-IR datasets Drone Vehicle,VEDAI,and LLVIP.In addition,GLFDet is more lightweight than other comparable models,making it more appealing to edge devices such as drones.The code is available at https://github.com/lao chen330/GLFDet.展开更多
Underwater pipeline inspection plays a vital role in the proactive maintenance and management of critical marine infrastructure and subaquatic systems.However,the inspection of underwater pipelines presents a challeng...Underwater pipeline inspection plays a vital role in the proactive maintenance and management of critical marine infrastructure and subaquatic systems.However,the inspection of underwater pipelines presents a challenge due to factors such as light scattering,absorption,restricted visibility,and ambient noise.The advancement of deep learning has introduced powerful techniques for processing large amounts of unstructured and imperfect data collected from underwater environments.This study evaluated the efficacy of the You Only Look Once(YOLO)algorithm,a real-time object detection and localization model based on convolutional neural networks,in identifying and classifying various types of pipeline defects in underwater settings.YOLOv8,the latest evolution in the YOLO family,integrates advanced capabilities,such as anchor-free detection,a cross-stage partial network backbone for efficient feature extraction,and a feature pyramid network+path aggregation network neck for robust multi-scale object detection,which make it particularly well-suited for complex underwater environments.Due to the lack of suitable open-access datasets for underwater pipeline defects,a custom dataset was captured using a remotely operated vehicle in a controlled environment.This application has the following assets available for use.Extensive experimentation demonstrated that YOLOv8 X-Large consistently outperformed other models in terms of pipe defect detection and classification and achieved a strong balance between precision and recall in identifying pipeline cracks,rust,corners,defective welds,flanges,tapes,and holes.This research establishes the baseline performance of YOLOv8 for underwater defect detection and showcases its potential to enhance the reliability and efficiency of pipeline inspection tasks in challenging underwater environments.展开更多
【Objective】This study aimed to establish a quintuple PCR method for rapid and simultaneous detection of Ralstonia solanacearum,Fusarium spp.,Pectobacterium spp.,Enterobacter spp.,and Pythium spp.,which provided tech...【Objective】This study aimed to establish a quintuple PCR method for rapid and simultaneous detection of Ralstonia solanacearum,Fusarium spp.,Pectobacterium spp.,Enterobacter spp.,and Pythium spp.,which provided technical support for early diagnosis of various soil-borne diseases on ginger.【Method】For five types of soil-borne pathogens causing ginger bacterial wilt and rhizome rot,specific primer combinations were designed and screened,the optimal quintuple reaction system was established by exploring optimal primer concentrations,annealing temperature,and sensitivity,and was applied to detect field plant samples to verify its utility.【Result】Specific primers pairs Rs1F/Rs1R,En1F/En1R,and Py1F/Py1R were designed according to flic gene of Ralstonia solanacearum,rpoB gene of Enterobacter spp.,and 18S rDNA of Pythium spp.,and combined with reported Fusarium spp.specific primers Fu3/Fu4 and specific primers 23SPecF/23SPecR of Pectobacterium spp.,a quintuple PCR reaction system for ginger soil-borne pathogens has been established(25.00μL):above primer dosage was 1.20,0.20,0.60,1.60,and 0.15μL respectively;2×PCR Mix 12.50μL;DNA templates of different pathogens were 1.00μL each;added ddH_(2)O to 25.00μL.Annealing temperature was optimized to 55.4℃.The specific fragments with sizes of 516,370,266,207,and 159 bp could be amplified simultaneously in the established quintuple PCR system,and the detection limit of this system for Ralstonia solanacearum,Enterobacter spp.and Pythium spp.reached 10^(-1)pg/μL,for Fusarium spp.and Pectobacterium spp.was 1 pg/μL,and for detecting five pathogens simultaneously was 10^(3)pg/μL.The multiplex PCR system established in this study could successfully detect the diseased plant samples from the field.【Conclusion】The quintuple PCR system established is able to rapid ly and accurately detect Ralstonia solanacearum,Fusarium spp.,Pectobacterium spp.,Enterobacter spp.,and Pythium spp.,which provides a useful tool for timely diagnosis and epidemic monitoring of various soil-borne diseases of ginger.展开更多
In this study,a multifunctional aptamer-conjugated magnetic covalent organic framework(COF)-CuO/Au nanozyme(MCOF-CuO/Au@apt)was developed as a“three-in-one”platform for dual-signal colorimetric and fluorescent detec...In this study,a multifunctional aptamer-conjugated magnetic covalent organic framework(COF)-CuO/Au nanozyme(MCOF-CuO/Au@apt)was developed as a“three-in-one”platform for dual-signal colorimetric and fluorescent detection of Vibrio parahaemolyticus.The nanozyme integrated magnetic separation,peroxidase-like catalytic activity,and specific target recognition through an aptamer-based strategy.Upon binding to V.parahaemolyticus,the catalytic oxidation of tetra-aminophenylethylene(TPE-4A)by the nanozyme was selectively inhibited,resulting in distinct colorimetric and fluorescent signals that significantly enhanced the detection accuracy and reliability.The proposed method exhibited high sensitivity,with limits of detection(LOD)of 21 and 7 CFU/mL for the colorimetric and fluorescent assays,respectively.The performance of this method was validated using real seafood samples,including Penaeus vannamei,Mytilus coruscus,and Crassostrea gigas,which showed high recovery rates(101.11%-107.30%)and excellent reproducibility.The system also demonstrated strong specificity and accuracy under various conditions,confirming its robustness and practical applicability.Collectively,this innovative platform presents a promising solution for the rapid,versatile,and sensitive detection of V.parahaemolyticus in seafood,with considerable potential to advance food safety diagnosis and on-site monitoring.展开更多
Diesel accounts for over 60%of the products derived from direct coal liquefaction(DCL).Compared to petroleum-based diesel,DCL diesel exhibits a cetane number ranging from 30 to 40,which fails to meet the automotive di...Diesel accounts for over 60%of the products derived from direct coal liquefaction(DCL).Compared to petroleum-based diesel,DCL diesel exhibits a cetane number ranging from 30 to 40,which fails to meet the automotive diesel standard requirement of≥45.Therefore,rapid and accurate analysis of its chemical composition is crucial for property optimization to meet fuel specifications by component blending.Thought traditional methods like gas chromatography offer high accuracy,they are unsuitable for rapid online analysis under industrial conditions.Near-infrared(NIR)spectroscopy can provide advantages in rapid,non-destructive analysis.Its application however,is limited by the complexity of spectral data interpretation.Machine learning(ML)is effective method for extracting valuable information from spectra and establishing high-precision prediction models.This study integrates NIR spectroscopy with ML to construct a spectral-composition database for DCL diesel.Feature extraction was performed using the correlation coefficient and mutual information methods to screen key wavelength variables and reduce data dimensionality.Subsequently,the predictive performance of three ML models—Lasso,SVR and XGBoost—was compared.Results indicate that excluding spectral data with absorbance greater than 1 significantly enhances model accuracy,increasing the test set R^(2) from 0.85 to 0.96.After feature extraction,the optimal number of wavelength variables was reduced to 177,substantially improving computational efficiency.Among the models evaluated,the SVR-MI-0.9 model,based on mutual information feature selection,demonstrated the best performance,achieving training and test set R^(2) values both exceeding 0.98.This model enables precise prediction of paraffin,naphthene,and aromatic hydrocarbon contents.This research provides a robust methodology for intelligent online quality monitoring.An intelligent NIR spectroscopy data analysis software was independently developed based on the established model.Compared with comprehensive two-dimensional gas chromatography,the software reduced the analysis time by over 98%,with an absolute prediction error below 0.2%.Thus,rapid analysis of DCL diesel components was successfully realized.展开更多
In the field of smart agriculture,accurate and efficient object detection technology is crucial for automated crop management.A particularly challenging task in this domain is small object detection,such as the identi...In the field of smart agriculture,accurate and efficient object detection technology is crucial for automated crop management.A particularly challenging task in this domain is small object detection,such as the identification of immature fruits or early stage disease spots.These objects pose significant difficulties due to their small pixel coverage,limited feature information,substantial scale variations,and high susceptibility to complex background interference.These challenges frequently result in inadequate accuracy and robustness in current detection models.This study addresses two critical needs in the cashew cultivation industry—fruitmaturity and anthracnose detection—by proposing an improved YOLOv11-NSDDil model.The method introduces three key technological innovations:(1)The SDDil module is designed and integrated into the backbone network.This module combines depthwise separable convolution with the SimAM attention mechanism to expand the receptive field and enhance contextual semantic capture at a low computational cost,effectively alleviating the feature deficiency problem caused by limited pixel coverage of small objects.Simultaneously,the SDmodule dynamically enhances discriminative features and suppresses background noise,significantly improving the model’s feature discrimination capability in complex environments;(2)The introduction of the DynamicScalSeq-Zoom_cat neck network,significantly improving multi-scale feature fusion;and(3)The optimization of the Minimum Point Distance Intersection over Union(MPDIoU)loss function,which enhances bounding box localization accuracy byminimizing vertex distance.Experimental results on a self-constructed cashew dataset containing 1123 images demonstrate significant performance improvements in the enhanced model:mAP50 reaches 0.825,a 4.6% increase compared to the originalYOLOv11;mAP50-95 improves to 0.624,a 6.5% increase;and recall rises to 0.777,a 2.4%increase.This provides a reliable technical solution for intelligent quality inspection of agricultural products and holds broad application prospects.展开更多
Dear Editor,This letter studies the motion planning issue for an autonomous underwater vehicle(AUV)in obstacle environment.We propose a novel integrated detection-communication waveform that enables simultaneous obsta...Dear Editor,This letter studies the motion planning issue for an autonomous underwater vehicle(AUV)in obstacle environment.We propose a novel integrated detection-communication waveform that enables simultaneous obstacle detection and self-localization.展开更多
Zero-day attacks present a critical cybersecurity challenge for Internet of things(IoT)infrastructures,where the inability of signature-based intrusion detection systems(IDSs)to recognize novel threat behaviors compro...Zero-day attacks present a critical cybersecurity challenge for Internet of things(IoT)infrastructures,where the inability of signature-based intrusion detection systems(IDSs)to recognize novel threat behaviors compromises both system reliability and operational continuity.Existing hybrid IDS solutions often struggle to balance accurate classification of known attacks with reliable anomaly detection,particularly under the computational constraints of IoT environments.To address this gap,we introduce ZeroDefense,an adaptive fusion-based IDS designed for simultaneous detection of known intrusions and emerging zero-day threats.The framework employs a four-layer architecture consisting of i)feature standardization and class balancing,ii)anomaly detection using isolation forest,autoencoder,and local outlier factor,iii)fine-grained attack classification via random forest,extreme gradient boosting(XGBoost),light gradient boosting machine(LightGBM),and attentive interpretable tabular learning(TabNet),and iv)a confidence-aware fusion engine that adaptively selects the most reliable decision path.Suspicious or previously unseen traffic is isolated early through fused anomaly scoring,while benign and known-malicious flows are processed through supervised classification for precise attack labeling.With an anomaly cascaded decision pipeline,a dynamic confidence-driven fusion mechanism,and a deploymentconscious design,ZeroDefense enables real-time inference on IoT edge gateways.Evaluation on the CICIoT2023 benchmark demonstrates 99.94% overall accuracy and 95.64%macro-average F1-score for known attacks,while 5.76% of traffic is successfully flagged as potential zero-day activity,with inference latency maintained below 100 ms/flow.These results indicate that ZeroDefense offers a scalable,resilient,and practically deployable defense capability for modern IoT infrastructures.展开更多
Distributed Denial-of-Service(DDoS)attacks pose severe threats to Industrial Control Networks(ICNs),where service disruption can cause significant economic losses and operational risks.Existing signature-based methods...Distributed Denial-of-Service(DDoS)attacks pose severe threats to Industrial Control Networks(ICNs),where service disruption can cause significant economic losses and operational risks.Existing signature-based methods are ineffective against novel attacks,and traditional machine learning models struggle to capture the complex temporal dependencies and dynamic traffic patterns inherent in ICN environments.To address these challenges,this study proposes a deep feature-driven hybrid framework that integrates Transformer,BiLSTM,and KNN to achieve accurate and robust DDoS detection.The Transformer component extracts global temporal dependencies from network traffic flows,while BiLSTM captures fine-grained sequential dynamics.The learned embeddings are then classified using an instance-based KNN layer,enhancing decision boundary precision.This cascaded architecture balances feature abstraction and locality preservation,improving both generalization and robustness.The proposed approach was evaluated on a newly collected real-time ICN traffic dataset and further validated using the public CIC-IDS2017 and Edge-IIoT datasets to demonstrate generalization.Comprehensive metrics including accuracy,precision,recall,F1-score,ROC-AUC,PR-AUC,false positive rate(FPR),and detection latency were employed.Results show that the hybrid framework achieves 98.42%accuracy with an ROC-AUC of 0.992 and FPR below 1%,outperforming baseline machine learning and deep learning models.Robustness experiments under Gaussian noise perturbations confirmed stable performance with less than 2%accuracy degradation.Moreover,detection latency remained below 2.1 ms per sample,indicating suitability for real-time ICS deployment.In summary,the proposed hybrid temporal learning and instance-based classification model offers a scalable and effective solution for DDoS detection in industrial control environments.By combining global contextual modeling,sequential learning,and instance-based refinement,the framework demonstrates strong adaptability across datasets and resilience against noise,providing practical utility for safeguarding critical infrastructure.展开更多
To address critical challenges in nighttime ship detection—high small-target missed detection(over 20%),insufficient lightweighting,and limited generalization due to scarce,low-quality datasets—this study proposes a...To address critical challenges in nighttime ship detection—high small-target missed detection(over 20%),insufficient lightweighting,and limited generalization due to scarce,low-quality datasets—this study proposes a systematic solution.First,a high-quality Night-Ships dataset is constructed via CycleGAN-based day-night transfer,combined with a dual-threshold cleaning strategy(Laplacian variance sharpness filtering and brightness-color deviation screening).Second,a Cross-stage Lightweight Fusion-You Only Look Once version 8(CLF-YOLOv8)is proposed with key improvements:the Neck network is reconstructed by replacing Cross Stage Partial(CSP)structure with the Cross Stage Partial Multi-Scale Convolutional Block(CSP-MSCB)and integrating Bidirectional Feature Pyramid Network(BiFPN)for weighted multi-scale fusion to enhance small-target detection;a Lightweight Shared Convolutional and Separated Batch Normalization Detection-Head(LSCSBD-Head)with shared convolutions and layer-wise Batch Normalization(BN)reduces parameters to 1.8M(42% fewer than YOLOv8n);and the FocalMinimum Point Distance Intersection over Union(Focal-MPDIoU)loss combines Minimum Point Distance Intersection over Union(MPDIoU)geometric constraints and Focal weighting to optimize low-overlap targets.Experiments show CLFYOLOv8 achieves 97.6%mAP@0.5(0.7% higher than YOLOv8n)with 1.8 M parameters,outperforming mainstream models in small-target detection,overlapping target discrimination,and adaptability to complex lighting.展开更多
Small object detection has been a focus of attention since the emergence of deep learning-based object detection.Although classical object detection frameworks have made significant contributions to the development of...Small object detection has been a focus of attention since the emergence of deep learning-based object detection.Although classical object detection frameworks have made significant contributions to the development of object detection,there are still many issues to be resolved in detecting small objects due to the inherent complexity and diversity of real-world visual scenes.In particular,the YOLO(You Only Look Once)series of detection models,renowned for their real-time performance,have undergone numerous adaptations aimed at improving the detection of small targets.In this survey,we summarize the state-of-the-art YOLO-based small object detection methods.This review presents a systematic categorization of YOLO-based approaches for small-object detection,organized into four methodological avenues,namely attention-based feature enhancement,detection-head optimization,loss function,and multi-scale feature fusion strategies.We then examine the principal challenges addressed by each category.Finally,we analyze the performance of thesemethods on public benchmarks and,by comparing current approaches,identify limitations and outline directions for future research.展开更多
Breast cancer screening programs rely heavily on mammography for early detection;however,diagnostic performance is strongly affected by inter-reader variability,breast density,and the limitations of conven-tional comp...Breast cancer screening programs rely heavily on mammography for early detection;however,diagnostic performance is strongly affected by inter-reader variability,breast density,and the limitations of conven-tional computer-aided detection systems.Recent advances in deep learning have enabled more robust and scalable solutions for large-scale screening,yet a systematic comparison of modern object detection architectures on nationally representative datasets remains limited.This study presents a comprehensive quantitative comparison of prominent deep learning–based object detection architectures for Artificial Intelligence-assisted mammography analysis using the MammosighTR dataset,developed within the Turkish National Breast Cancer Screening Program.The dataset comprises 12,740 patient cases collected between 2016 and 2022,annotated with BI-RADS categories,breast density levels,and lesion localization labels.A total of 31 models were evaluated,including One-Stage,Two-Stage,and Transformer-based architectures,under a unified experimental framework at both patient and breast levels.The results demonstrate that Two-Stage architectures consistently outperform One-Stage models,achieving approximately 2%–4%higher Macro F1-Scores and more balanced precision–recall trade-offs,with Double-Head R-CNN and Dynamic R-CNN yielding the highest overall performance(Macro F1≈0.84–0.86).This advantage is primarily attributed to the region proposal mechanism and improved class balance inherent to Two-Stage designs.One-Stage detectors exhibited higher sensitivity and faster inference,reaching Recall values above 0.88,but experienced minor reductions in Precision and overall accuracy(≈1%–2%)compared with Two-Stage models.Among Transformer-based architectures,Deformable DEtection TRansformer demonstrated strong robustness and consistency across datasets,achieving Macro F1-Scores comparable to CNN-based detectors(≈0.83–0.85)while exhibiting minimal performance degradation under distributional shifts.Breast density–based analysis revealed increased misclassification rates in medium-density categories(types B and C),whereas Transformer-based architectures maintained more stable performance in high-density type D tissue.These findings quantitatively confirm that both architectural design and tissue characteristics play a decisive role in diagnostic accuracy.Overall,the study provides a reproducible benchmark and highlights the potential of hybrid approaches that combine the accuracy of Two-Stage detectors with the contextual modeling capability of Transformer architectures for clinically reliable breast cancer screening systems.展开更多
文摘Background:The reputation system has been designed as an effective mechanism to reduce risks associated with online shopping for customers.However,it is vulnerable to rating fraud.Some raters may inject unfairly high or low ratings to the system so as to promote their own products or demote their competitors.Method:This study explores the rating fraud by differentiating the subjective fraud from objective fraud.Then it discusses the effectiveness of blockchain technology in objective fraud and its limitation in subjective fraud,especially the rating fraud.Lastly,it systematically analyzes the robustness of blockchain-based reputation systems in each type of rating fraud.Results:The detection of fraudulent raters is not easy since they can behave strategically to camouflage themselves.We explore the potential strengths and limitations of blockchain-based reputation systems under two attack goals:ballot-stuffing and bad-mouthing,and various attack models including constant attack,camouflage attack,whitewashing attack and sybil attack.Blockchain-based reputation systems are more robust against bad-mouthing than ballot-stuffing fraud.Conclusions:Blockchain technology provides new opportunities for redesigning the reputation system.Blockchain systems are very effective in preventing objective information fraud,such as loan application fraud,where fraudulent information is fact-based.However,their effectiveness is limited in subjective information fraud,such as rating fraud,where the ground-truth is not easily validated.Blockchain systems are effective in preventing bad mouthing and whitewashing attack,but they are limited in detecting ballot-stuffing under sybil attack,constant attacks and camouflage attack.
基金Supported by the National Natural Science Foun-dation of China (60006002) Natural Science Research Project of Education Department of Guangdong Province of China (02019)
文摘With the complexity of integrated circuits is continually increasing, a local defect in circuits may cause multiple faults. The behavior of a digital circuit with a multiple fault may significantly differ from that of a single fault. A new method for the detection of multiple faults in digital circuits is presented in this paper, the method is based on binary decision diagram （BDD）. First of all, the BDDs for the normal circuit and faulty circuit are built respectively. Secondly, a test BDD is obtained by the XOR operation of the BDDs corresponds to normal circuit and faulty circuit. In the test BDD, each input assignment that leads to the leaf node labeled 1 is a test vector of multiple faults. Therefore, the test set of multiple faults is generated by searching for the type of input assignments in the test BDD. Experimental results on some digital circuits show the feasibility of the approach presented in this paper.
基金the International Science and Technology Cooperation Project of the Shenzhen Science and Technology Commission(GJHZ20200731095804014).
文摘This study designs a microstrip patch antenna with an inverted T-type notch in the partial ground to detect tumorcells inside the human breast.The size of the current antenna is small enough(18mm×21mm×1.6mm)todistribute around the breast phantom.The operating frequency has been observed from6–14GHzwith a minimumreturn loss of−61.18 dB and themaximumgain of current proposed antenna is 5.8 dBiwhich is flexiblewith respectto the size of antenna.After the distribution of eight antennas around the breast phantom,the return loss curveswere observed in the presence and absence of tumor cells inside the breast phantom,and these observations showa sharp difference between the presence and absence of tumor cells.The simulated results show that this proposedantenna is suitable for early detection of cancerous cells inside the breast.
文摘Surface-enhanced Raman scattering(SERS)is a powerful technology for obtaining vibrational information from molecules that present in different chemical or biological environments.This paper presents a 3D SERS substrate based on nanocone forests.The substrates are prepared by using plasma treatment technique,which is a simple,fast and high-throughput approach.The SERS substrate based on nanocone forests exhibits high sensitivity.In the experiment,miRNA with a concentration as low as 10-10 M can be achieved.Meanwhile,the proposed SERS substrate shows a high uniformity over a large area.These experimental results demonstrate great potential of the 3D SERS substrate in wide applications.
基金the National Natural Science Foundation of China(No60432040)the Guangxi Key Laboratory Foundation(No,063006-5G)
文摘To solve the problem that the conventional detections in DS-CDMA suffer from high complexity and poor robustness for the time-hopping pulse signals, the received pulse signals were remodeled, and a mulfipath-free detection scheme, which provides a simple approach to select samples of received signals, was introduced. By this scheme, the subsequent multiuser detection （MUD） would get rid of the mis- match due to the correlative multipath signal in IR-UWB. In addition, a computationally efficient recur-sive least squares （RLS） type algorithm based on least mean fourth （LMF） criterion is derived to suppress multi-access interference. The proposed multiuser detection algorithm performs well at low complexity, even in dense muhipath environment.
基金supported by the National Key Research and Development Program of China (No.2024YFA1509600)National Natural Science Foundation of China (Nos.22474010 and 22274012)the Fundamental Research Funds for the Central Universities (No.2233300007)。
文摘Ferroptosis has exhibited great potential in therapies and intracellular reducing agents of sulfur species(RSSs) in the thiol-dependent redox systems are crucial in ferroptosis.This makes the simultaneous detection of multiple RSSs significant for evaluating ferroptosis therapy.However,the traditional techniques,including fluorescent(FL) imaging and electrospray ionization-based mass spectrometry(MS) detection,cannot achieve the discrimination of different RSSs.Herein,simultaneous MS detection of multiple RSSs,including cysteine(Cys),homocysteine(Hcy),glutathione(GSH) and hydrogen sulfide(H_(2)S),was obtained upon enhancing ionization efficiency by a fluorescent probe(NBD-O-1).Based on the interaction between NBD-O-1 and RSSs,the complex of RSSs with a fragment of NBD-O-1 can be generated,which can be easily ionized for MS detection in the negative mode.Therefore,the intracellular RSSs can be well detected upon the incubation of He La cells with the probe of NBD-O-1,exhibiting the total RSS levels by the FL imaging and further providing expression of each RSS by enhanced MS detection.Furthermore,the RSSs during ferroptosis in He La cells have been evaluated using the present strategy,demonstrating the potential for ferroptosis examinations.This work has made an unconventional application of a fluorescent probe to enhance the detection of multiple RSSs by MS,providing significant molecular information for addressing the ferroptosis mechanism.
基金supported by the National Natural Science Foundation of China(No.62276204)the Fundamental Research Funds for the Central Universities,China(No.YJSJ24011)+1 种基金the Natural Science Basic Research Program of Shaanxi,China(Nos.2022JM-340 and 2023-JC-QN-0710)the China Postdoctoral Science Foundation(Nos.2020T130494 and 2018M633470)。
文摘Visible and infrared(RGB-IR)fusion object detection plays an important role in security,disaster relief,etc.In recent years,deep-learning-based RGB-IR fusion detection methods have been developing rapidly,but still struggle to deal with the complex and changing scenarios captured by drones,mainly due to two reasons:(A)RGB-IR fusion detectors are susceptible to inferior inputs that degrade performance and stability.(B)RGB-IR fusion detectors are susceptible to redundant features that reduce accuracy and efficiency.In this paper,an innovative RGB-IR fusion detection framework based on global-local feature optimization,named GLFDet,is proposed to improve the detection performance and efficiency of drone-captured objects.The key components of GLFDet include a Global Feature Optimization(GFO)module,a Local Feature Optimization(LFO)module and a Channel Separation Fusion(CSF)module.Specifically,GFO calculates the information content of the input image from the frequency domain and optimizes the features holistically.Then,LFO dynamically selects high-value features and filters out low-value features before fusion,which significantly improves the efficiency of fusion.Finally,CSF fuses the RGB and IR features across the corresponding channels,which avoids the rearrangement of the channel relationships and enhances the model stability.Extensive experimental results show that the proposed method achieves the best performance on three popular RGB-IR datasets Drone Vehicle,VEDAI,and LLVIP.In addition,GLFDet is more lightweight than other comparable models,making it more appealing to edge devices such as drones.The code is available at https://github.com/lao chen330/GLFDet.
文摘Underwater pipeline inspection plays a vital role in the proactive maintenance and management of critical marine infrastructure and subaquatic systems.However,the inspection of underwater pipelines presents a challenge due to factors such as light scattering,absorption,restricted visibility,and ambient noise.The advancement of deep learning has introduced powerful techniques for processing large amounts of unstructured and imperfect data collected from underwater environments.This study evaluated the efficacy of the You Only Look Once(YOLO)algorithm,a real-time object detection and localization model based on convolutional neural networks,in identifying and classifying various types of pipeline defects in underwater settings.YOLOv8,the latest evolution in the YOLO family,integrates advanced capabilities,such as anchor-free detection,a cross-stage partial network backbone for efficient feature extraction,and a feature pyramid network+path aggregation network neck for robust multi-scale object detection,which make it particularly well-suited for complex underwater environments.Due to the lack of suitable open-access datasets for underwater pipeline defects,a custom dataset was captured using a remotely operated vehicle in a controlled environment.This application has the following assets available for use.Extensive experimentation demonstrated that YOLOv8 X-Large consistently outperformed other models in terms of pipe defect detection and classification and achieved a strong balance between precision and recall in identifying pipeline cracks,rust,corners,defective welds,flanges,tapes,and holes.This research establishes the baseline performance of YOLOv8 for underwater defect detection and showcases its potential to enhance the reliability and efficiency of pipeline inspection tasks in challenging underwater environments.
基金National Natural Science Foundation of China(32270237)Guangxi Key Research and Development Plan Project(Guike AB21238002)Basic Scientific Research Project of Guangxi Academy of Agricultural Sciences(Guinongke 2024YP082)。
文摘【Objective】This study aimed to establish a quintuple PCR method for rapid and simultaneous detection of Ralstonia solanacearum,Fusarium spp.,Pectobacterium spp.,Enterobacter spp.,and Pythium spp.,which provided technical support for early diagnosis of various soil-borne diseases on ginger.【Method】For five types of soil-borne pathogens causing ginger bacterial wilt and rhizome rot,specific primer combinations were designed and screened,the optimal quintuple reaction system was established by exploring optimal primer concentrations,annealing temperature,and sensitivity,and was applied to detect field plant samples to verify its utility.【Result】Specific primers pairs Rs1F/Rs1R,En1F/En1R,and Py1F/Py1R were designed according to flic gene of Ralstonia solanacearum,rpoB gene of Enterobacter spp.,and 18S rDNA of Pythium spp.,and combined with reported Fusarium spp.specific primers Fu3/Fu4 and specific primers 23SPecF/23SPecR of Pectobacterium spp.,a quintuple PCR reaction system for ginger soil-borne pathogens has been established(25.00μL):above primer dosage was 1.20,0.20,0.60,1.60,and 0.15μL respectively;2×PCR Mix 12.50μL;DNA templates of different pathogens were 1.00μL each;added ddH_(2)O to 25.00μL.Annealing temperature was optimized to 55.4℃.The specific fragments with sizes of 516,370,266,207,and 159 bp could be amplified simultaneously in the established quintuple PCR system,and the detection limit of this system for Ralstonia solanacearum,Enterobacter spp.and Pythium spp.reached 10^(-1)pg/μL,for Fusarium spp.and Pectobacterium spp.was 1 pg/μL,and for detecting five pathogens simultaneously was 10^(3)pg/μL.The multiplex PCR system established in this study could successfully detect the diseased plant samples from the field.【Conclusion】The quintuple PCR system established is able to rapid ly and accurately detect Ralstonia solanacearum,Fusarium spp.,Pectobacterium spp.,Enterobacter spp.,and Pythium spp.,which provides a useful tool for timely diagnosis and epidemic monitoring of various soil-borne diseases of ginger.
文摘In this study,a multifunctional aptamer-conjugated magnetic covalent organic framework(COF)-CuO/Au nanozyme(MCOF-CuO/Au@apt)was developed as a“three-in-one”platform for dual-signal colorimetric and fluorescent detection of Vibrio parahaemolyticus.The nanozyme integrated magnetic separation,peroxidase-like catalytic activity,and specific target recognition through an aptamer-based strategy.Upon binding to V.parahaemolyticus,the catalytic oxidation of tetra-aminophenylethylene(TPE-4A)by the nanozyme was selectively inhibited,resulting in distinct colorimetric and fluorescent signals that significantly enhanced the detection accuracy and reliability.The proposed method exhibited high sensitivity,with limits of detection(LOD)of 21 and 7 CFU/mL for the colorimetric and fluorescent assays,respectively.The performance of this method was validated using real seafood samples,including Penaeus vannamei,Mytilus coruscus,and Crassostrea gigas,which showed high recovery rates(101.11%-107.30%)and excellent reproducibility.The system also demonstrated strong specificity and accuracy under various conditions,confirming its robustness and practical applicability.Collectively,this innovative platform presents a promising solution for the rapid,versatile,and sensitive detection of V.parahaemolyticus in seafood,with considerable potential to advance food safety diagnosis and on-site monitoring.
基金Supported by National Natural Science Foundation of China(U24B6018,22178243)。
文摘Diesel accounts for over 60%of the products derived from direct coal liquefaction(DCL).Compared to petroleum-based diesel,DCL diesel exhibits a cetane number ranging from 30 to 40,which fails to meet the automotive diesel standard requirement of≥45.Therefore,rapid and accurate analysis of its chemical composition is crucial for property optimization to meet fuel specifications by component blending.Thought traditional methods like gas chromatography offer high accuracy,they are unsuitable for rapid online analysis under industrial conditions.Near-infrared(NIR)spectroscopy can provide advantages in rapid,non-destructive analysis.Its application however,is limited by the complexity of spectral data interpretation.Machine learning(ML)is effective method for extracting valuable information from spectra and establishing high-precision prediction models.This study integrates NIR spectroscopy with ML to construct a spectral-composition database for DCL diesel.Feature extraction was performed using the correlation coefficient and mutual information methods to screen key wavelength variables and reduce data dimensionality.Subsequently,the predictive performance of three ML models—Lasso,SVR and XGBoost—was compared.Results indicate that excluding spectral data with absorbance greater than 1 significantly enhances model accuracy,increasing the test set R^(2) from 0.85 to 0.96.After feature extraction,the optimal number of wavelength variables was reduced to 177,substantially improving computational efficiency.Among the models evaluated,the SVR-MI-0.9 model,based on mutual information feature selection,demonstrated the best performance,achieving training and test set R^(2) values both exceeding 0.98.This model enables precise prediction of paraffin,naphthene,and aromatic hydrocarbon contents.This research provides a robust methodology for intelligent online quality monitoring.An intelligent NIR spectroscopy data analysis software was independently developed based on the established model.Compared with comprehensive two-dimensional gas chromatography,the software reduced the analysis time by over 98%,with an absolute prediction error below 0.2%.Thus,rapid analysis of DCL diesel components was successfully realized.
基金supported by Hebei North University Doctoral Research Fund Project(No.BSJJ202315)the Youth Research Fund Project of Higher Education Institutions in Hebei Province(No.QN2024146).
文摘In the field of smart agriculture,accurate and efficient object detection technology is crucial for automated crop management.A particularly challenging task in this domain is small object detection,such as the identification of immature fruits or early stage disease spots.These objects pose significant difficulties due to their small pixel coverage,limited feature information,substantial scale variations,and high susceptibility to complex background interference.These challenges frequently result in inadequate accuracy and robustness in current detection models.This study addresses two critical needs in the cashew cultivation industry—fruitmaturity and anthracnose detection—by proposing an improved YOLOv11-NSDDil model.The method introduces three key technological innovations:(1)The SDDil module is designed and integrated into the backbone network.This module combines depthwise separable convolution with the SimAM attention mechanism to expand the receptive field and enhance contextual semantic capture at a low computational cost,effectively alleviating the feature deficiency problem caused by limited pixel coverage of small objects.Simultaneously,the SDmodule dynamically enhances discriminative features and suppresses background noise,significantly improving the model’s feature discrimination capability in complex environments;(2)The introduction of the DynamicScalSeq-Zoom_cat neck network,significantly improving multi-scale feature fusion;and(3)The optimization of the Minimum Point Distance Intersection over Union(MPDIoU)loss function,which enhances bounding box localization accuracy byminimizing vertex distance.Experimental results on a self-constructed cashew dataset containing 1123 images demonstrate significant performance improvements in the enhanced model:mAP50 reaches 0.825,a 4.6% increase compared to the originalYOLOv11;mAP50-95 improves to 0.624,a 6.5% increase;and recall rises to 0.777,a 2.4%increase.This provides a reliable technical solution for intelligent quality inspection of agricultural products and holds broad application prospects.
基金supported in part by the National Natural Science Foundation of China(U25A20473,62222314)the YanZhao Young Scientist Project of Hebei Province(F2024203047)+2 种基金the Natural Science Foundation of Hebei Province(F2022203001,F2024203072)the State Key Laboratory of Submarine Geoscience(sglkt2025-7)the Education Department Foundation of Hebei Province(JCZX2025027)。
文摘Dear Editor,This letter studies the motion planning issue for an autonomous underwater vehicle(AUV)in obstacle environment.We propose a novel integrated detection-communication waveform that enables simultaneous obstacle detection and self-localization.
文摘Zero-day attacks present a critical cybersecurity challenge for Internet of things(IoT)infrastructures,where the inability of signature-based intrusion detection systems(IDSs)to recognize novel threat behaviors compromises both system reliability and operational continuity.Existing hybrid IDS solutions often struggle to balance accurate classification of known attacks with reliable anomaly detection,particularly under the computational constraints of IoT environments.To address this gap,we introduce ZeroDefense,an adaptive fusion-based IDS designed for simultaneous detection of known intrusions and emerging zero-day threats.The framework employs a four-layer architecture consisting of i)feature standardization and class balancing,ii)anomaly detection using isolation forest,autoencoder,and local outlier factor,iii)fine-grained attack classification via random forest,extreme gradient boosting(XGBoost),light gradient boosting machine(LightGBM),and attentive interpretable tabular learning(TabNet),and iv)a confidence-aware fusion engine that adaptively selects the most reliable decision path.Suspicious or previously unseen traffic is isolated early through fused anomaly scoring,while benign and known-malicious flows are processed through supervised classification for precise attack labeling.With an anomaly cascaded decision pipeline,a dynamic confidence-driven fusion mechanism,and a deploymentconscious design,ZeroDefense enables real-time inference on IoT edge gateways.Evaluation on the CICIoT2023 benchmark demonstrates 99.94% overall accuracy and 95.64%macro-average F1-score for known attacks,while 5.76% of traffic is successfully flagged as potential zero-day activity,with inference latency maintained below 100 ms/flow.These results indicate that ZeroDefense offers a scalable,resilient,and practically deployable defense capability for modern IoT infrastructures.
基金supported by the Extral High Voltage Power Transmission Company,China Southern Power Grid Co.,Ltd.
文摘Distributed Denial-of-Service(DDoS)attacks pose severe threats to Industrial Control Networks(ICNs),where service disruption can cause significant economic losses and operational risks.Existing signature-based methods are ineffective against novel attacks,and traditional machine learning models struggle to capture the complex temporal dependencies and dynamic traffic patterns inherent in ICN environments.To address these challenges,this study proposes a deep feature-driven hybrid framework that integrates Transformer,BiLSTM,and KNN to achieve accurate and robust DDoS detection.The Transformer component extracts global temporal dependencies from network traffic flows,while BiLSTM captures fine-grained sequential dynamics.The learned embeddings are then classified using an instance-based KNN layer,enhancing decision boundary precision.This cascaded architecture balances feature abstraction and locality preservation,improving both generalization and robustness.The proposed approach was evaluated on a newly collected real-time ICN traffic dataset and further validated using the public CIC-IDS2017 and Edge-IIoT datasets to demonstrate generalization.Comprehensive metrics including accuracy,precision,recall,F1-score,ROC-AUC,PR-AUC,false positive rate(FPR),and detection latency were employed.Results show that the hybrid framework achieves 98.42%accuracy with an ROC-AUC of 0.992 and FPR below 1%,outperforming baseline machine learning and deep learning models.Robustness experiments under Gaussian noise perturbations confirmed stable performance with less than 2%accuracy degradation.Moreover,detection latency remained below 2.1 ms per sample,indicating suitability for real-time ICS deployment.In summary,the proposed hybrid temporal learning and instance-based classification model offers a scalable and effective solution for DDoS detection in industrial control environments.By combining global contextual modeling,sequential learning,and instance-based refinement,the framework demonstrates strong adaptability across datasets and resilience against noise,providing practical utility for safeguarding critical infrastructure.
基金the Shandong Provincial Key Research and Development Program(Grant No.2024SFGC0201).
文摘To address critical challenges in nighttime ship detection—high small-target missed detection(over 20%),insufficient lightweighting,and limited generalization due to scarce,low-quality datasets—this study proposes a systematic solution.First,a high-quality Night-Ships dataset is constructed via CycleGAN-based day-night transfer,combined with a dual-threshold cleaning strategy(Laplacian variance sharpness filtering and brightness-color deviation screening).Second,a Cross-stage Lightweight Fusion-You Only Look Once version 8(CLF-YOLOv8)is proposed with key improvements:the Neck network is reconstructed by replacing Cross Stage Partial(CSP)structure with the Cross Stage Partial Multi-Scale Convolutional Block(CSP-MSCB)and integrating Bidirectional Feature Pyramid Network(BiFPN)for weighted multi-scale fusion to enhance small-target detection;a Lightweight Shared Convolutional and Separated Batch Normalization Detection-Head(LSCSBD-Head)with shared convolutions and layer-wise Batch Normalization(BN)reduces parameters to 1.8M(42% fewer than YOLOv8n);and the FocalMinimum Point Distance Intersection over Union(Focal-MPDIoU)loss combines Minimum Point Distance Intersection over Union(MPDIoU)geometric constraints and Focal weighting to optimize low-overlap targets.Experiments show CLFYOLOv8 achieves 97.6%mAP@0.5(0.7% higher than YOLOv8n)with 1.8 M parameters,outperforming mainstream models in small-target detection,overlapping target discrimination,and adaptability to complex lighting.
基金supported in part by the by Chongqing Research Program of Basic Research and Frontier Technology under Grant CSTB2025NSCQ-GPX1309.
文摘Small object detection has been a focus of attention since the emergence of deep learning-based object detection.Although classical object detection frameworks have made significant contributions to the development of object detection,there are still many issues to be resolved in detecting small objects due to the inherent complexity and diversity of real-world visual scenes.In particular,the YOLO(You Only Look Once)series of detection models,renowned for their real-time performance,have undergone numerous adaptations aimed at improving the detection of small targets.In this survey,we summarize the state-of-the-art YOLO-based small object detection methods.This review presents a systematic categorization of YOLO-based approaches for small-object detection,organized into four methodological avenues,namely attention-based feature enhancement,detection-head optimization,loss function,and multi-scale feature fusion strategies.We then examine the principal challenges addressed by each category.Finally,we analyze the performance of thesemethods on public benchmarks and,by comparing current approaches,identify limitations and outline directions for future research.
文摘Breast cancer screening programs rely heavily on mammography for early detection;however,diagnostic performance is strongly affected by inter-reader variability,breast density,and the limitations of conven-tional computer-aided detection systems.Recent advances in deep learning have enabled more robust and scalable solutions for large-scale screening,yet a systematic comparison of modern object detection architectures on nationally representative datasets remains limited.This study presents a comprehensive quantitative comparison of prominent deep learning–based object detection architectures for Artificial Intelligence-assisted mammography analysis using the MammosighTR dataset,developed within the Turkish National Breast Cancer Screening Program.The dataset comprises 12,740 patient cases collected between 2016 and 2022,annotated with BI-RADS categories,breast density levels,and lesion localization labels.A total of 31 models were evaluated,including One-Stage,Two-Stage,and Transformer-based architectures,under a unified experimental framework at both patient and breast levels.The results demonstrate that Two-Stage architectures consistently outperform One-Stage models,achieving approximately 2%–4%higher Macro F1-Scores and more balanced precision–recall trade-offs,with Double-Head R-CNN and Dynamic R-CNN yielding the highest overall performance(Macro F1≈0.84–0.86).This advantage is primarily attributed to the region proposal mechanism and improved class balance inherent to Two-Stage designs.One-Stage detectors exhibited higher sensitivity and faster inference,reaching Recall values above 0.88,but experienced minor reductions in Precision and overall accuracy(≈1%–2%)compared with Two-Stage models.Among Transformer-based architectures,Deformable DEtection TRansformer demonstrated strong robustness and consistency across datasets,achieving Macro F1-Scores comparable to CNN-based detectors(≈0.83–0.85)while exhibiting minimal performance degradation under distributional shifts.Breast density–based analysis revealed increased misclassification rates in medium-density categories(types B and C),whereas Transformer-based architectures maintained more stable performance in high-density type D tissue.These findings quantitatively confirm that both architectural design and tissue characteristics play a decisive role in diagnostic accuracy.Overall,the study provides a reproducible benchmark and highlights the potential of hybrid approaches that combine the accuracy of Two-Stage detectors with the contextual modeling capability of Transformer architectures for clinically reliable breast cancer screening systems.
