At an age when most teens are figuring out high school,Siddharth is already shaping the future of medical tech.The 14⁃year⁃old boy from Dallas has created an AI⁃powered app,Circadian AI,capable of detecting heart dise...At an age when most teens are figuring out high school,Siddharth is already shaping the future of medical tech.The 14⁃year⁃old boy from Dallas has created an AI⁃powered app,Circadian AI,capable of detecting heart disease in just 7 seconds using only a smartphone's microphone.展开更多
BACKGROUND Diagnosing bacterial infections(BI)in patients with cirrhosis can be challenging because of unclear symptoms,low diagnostic accuracy,and lengthy culture testing times.Various biomarkers have been studied,in...BACKGROUND Diagnosing bacterial infections(BI)in patients with cirrhosis can be challenging because of unclear symptoms,low diagnostic accuracy,and lengthy culture testing times.Various biomarkers have been studied,including serum procal-citonin(PCT)and presepsin.However,the diagnostic performance of these markers remains unclear,requiring further informative studies to ascertain their diagnostic value.AIM To evaluate the pooled diagnostic performance of PCT and presepsin in detecting BI among patients with cirrhosis.INTRODUCTION Bacterial infections(BI)commonly occur in patients with cirrhosis,resulting in poor outcomes,including the development of cirrhotic complications,septic shock,acute-on-chronic liver failure(ACLF),multiple organ failures,and mortality[1,2].BI is observed in 20%-30%of hospitalized patients,with and without ACLF[3].Patients with cirrhosis are susceptible to BI because of internal and external factors.The major internal factors are changes in gut microbial composition and function,bacterial translocation,and cirrhosis-associated immune dysfunction syndrome[4,5].External factors include alcohol use,proton-pump inhibitor use,frailty,readmission,and invasive procedures.Spontaneous bacterial peritonitis(SBP),urinary tract infection,pneumonia,and primary bacteremia are the common BIs in hospit-alized patients with cirrhosis[6].Early diagnosis and adequate empirical antibiotic therapy are two critical factors that improve the prognosis of BI in patients with cirrhosis.However,early detection of BI in cirrhosis is challenging due to subtle clinical signs and symptoms,low sensitivity and specificity of systemic inflammatory response syndrome criteria,and low sensitivity of bacterial cultures.Thus,effective biomarkers need to be identified for the early detection of BI.Several biomarkers have been evaluated,but their efficacy in detecting BI is unclear.Procalcitonin(PCT)is a precursor of the hormone calcitonin,which is secreted by parafollicular cells of the thyroid gland[7].In the presence of BI,PCT gene expression increases in extrathyroidal tissues,causing a subsequent increase in serum PCT level[8].Changes in serum PCT are detectable as early as 4 hours after infection onset and peaks between 8 and 24 hours,making it a valuable diagnostic biomarker for BI.Several studies have demonstrated the favorable diagnostic accuracy of PCT in the diagnosis of BI in individuals with cirrhosis[9-13]and without cirrhosis[14-16].Since 2014,two meta-analyses have been published on the diagnostic value of PCT for SBP and BI in patients with cirrhosis[17,18].Other related studies have been conducted since then[10-12,19-33].Serum presepsin has recently emerged as a promising biomarker for diagnosing BI.This biomarker is the N-terminal fraction protein of the soluble CD14 g-negative bacterial lipopolysaccharide–lipopolysaccharide binding protein(sCD14-LPS-LBP)complex,which is cleaved by inflammatory serum protease in response to BI[34].Presepsin levels increase within 2 hours and peaks in 3 hours[35].This is useful for detecting BI since presepsin levels increase earlier than serum Our systematic review and meta-analysis was performed with adherence to PRISMA guidelines[37].展开更多
Tomato plant diseases often first manifest on the leaves,making the detection of tomato leaf diseases particularly crucial for the tomato cultivation industry.However,conventional deep learning models face challenges ...Tomato plant diseases often first manifest on the leaves,making the detection of tomato leaf diseases particularly crucial for the tomato cultivation industry.However,conventional deep learning models face challenges such as large model sizes and slow detection speeds when deployed on resource-constrained platforms and agricultural machinery.This paper proposes a lightweight model for detecting tomato leaf diseases,named LT-YOLO,based on the YOLOv8n architecture.First,we enhance the C2f module into a RepViT Block(RVB)with decoupled token and channel mixers to reduce the cost of feature extraction.Next,we incorporate a novel Efficient Multi-Scale Attention(EMA)mechanism in the deeper layers of the backbone to improve detection of critical disease features.Additionally,we design a lightweight detection head,LT-Detect,using Partial Convolution(PConv)to significantly reduce the classification and localization costs during detection.Finally,we introduce a Receptive Field Block(RFB)in the shallow layers of the backbone to expand the model’s receptive field,enabling effective detection of diseases at various scales.The improved model reduces the number of parameters by 43%and the computational load by 50%.Additionally,it achieves a mean Average Precision(mAP)of 90.9%on a publicly available dataset containing 3641 images of tomato leaf diseases,with only a 0.7%decrease compared to the baseline model.This demonstrates that the model maintains excellent accuracy while being lightweight,making it suitable for rapid detection of tomato leaf diseases.展开更多
Driver distraction is a leading cause of traffic accidents,with fatigue being a significant contributor.This paper introduces a novel method for detecting driver distraction by analyzing facial features using machine ...Driver distraction is a leading cause of traffic accidents,with fatigue being a significant contributor.This paper introduces a novel method for detecting driver distraction by analyzing facial features using machine deep learning and 68 face model.The proposed system assesses driver tiredness by measuring the distance between key facial landmarks,such as the distance between the eyes and the angle of the mouth,to evaluate signs of drowsiness or disengagement.Real-time video feed analysis allows for continuous monitoring of the driver’s face,enabling the system to detect behavioral cues associated with distraction,such as eye closures or changes in facial expressions.The effectiveness of this method is demonstrated through a series of experiments on a dataset of driver videos,which proves that the approach can accurately assess tiredness and distraction levels under various driving conditions.By focusing on facial landmarks,the system is computationally efficient and capable of operating in real-time,making it a practical solution for in-vehicle safety systems.This paper discusses the system’s performance,limitations,and potential for future enhancements,including integration with other in-vehicle technologies to provide comprehensive driver monitoring.展开更多
Detection of ore waste is crucial for achieving automation in mineral metallurgy production.However,deep learning-based target detection algorithms still face several challenges in iron waste screening,including poor ...Detection of ore waste is crucial for achieving automation in mineral metallurgy production.However,deep learning-based target detection algorithms still face several challenges in iron waste screening,including poor lighting conditions in underground mining environments,dust disturbances,platform vibrations during operation,and limited resources for large-scale computing equipment.These factors contribute to extended computation times and unsatisfactory detection accuracy.To address these challenges,this paper proposes an ore waste detection algorithm based on an improved version of YOLOv5.To enhance feature extraction capabilities,the RepLKNet module is incorporated into the YOLOv5 backbone and neck networks.This module enhances the deformation information of feature extraction with the maximum effective Receptive Field to increase the model's accuracy.The Normalizationbased Attention Module(NAM)was introduced to enhance the attention mechanism by focusing on the most relevant features.This improves accuracy in detecting objects against noisy or unclear backgrounds,thereby further enhancing detection performance while reducing model parameters.Additionally,the loss function is optimized to constrain angular deviation using the SIOU loss function,which prevents the training frame from drifting during training and enhances convergence speed.To validate the performance of the proposed method,we tested it using a self-constructed dataset comprising 1,328 images obtained from the crushing station at Jinchuan Group's No.2 mine.The results indicate that,compared to YOLOv5s on the self-constructed dataset,the proposed algorithm achieves an 18.3%improvement in mAP(0.5),a 54%reduction in FLOPs,and a 52.53%decrease in model parameters.The effectiveness and superiority of the proposed algorithm are demonstrated through case studies and comparative analyses.展开更多
Depression is increasingly prevalent among adolescents and can profoundly impact their lives.However,the early detection of depression is often hindered by the timeconsuming diagnostic process and the absence of objec...Depression is increasingly prevalent among adolescents and can profoundly impact their lives.However,the early detection of depression is often hindered by the timeconsuming diagnostic process and the absence of objective biomarkers.In this study,we propose a novel approach for depression detection based on an affective brain-computer interface(aBCI)and the resting-state electroencephalogram(EEG).By fusing EEG features associated with both emotional and resting states,our method captures comprehensive depression-related information.The final depression detection model,derived through decision fusion with multiple independent models,further enhances detection efficacy.Our experiments involved 40 adolescents with depression and 40 matched controls.The proposed model achieved an accuracy of 86.54%on cross-validation and 88.20%on the independent test set,demonstrating the efficiency of multi-modal fusion.In addition,further analysis revealed distinct brain activity patterns between the two groups across different modalities.These findings hold promise for new directions in depression detection and intervention.展开更多
Detection of Autism Spectrum Disorder(ASD)is a crucial area of research,representing a foundational aspect of psychological studies.The advancement of technology and the widespread adoption of machine learning methodo...Detection of Autism Spectrum Disorder(ASD)is a crucial area of research,representing a foundational aspect of psychological studies.The advancement of technology and the widespread adoption of machine learning methodologies have brought significant attention to this field in recent years.Interdisciplinary efforts have further propelled research into detection methods.Consequently,this study aims to contribute to both the fields of psychology and computer science.Specifically,the goal is to apply machine learning techniques to limited data for the detection of Autism Spectrum Disorder.This study is structured into two distinct phases:data preprocessing and classification.In the data preprocessing phase,four datasets—Toddler,Children,Adolescent,and Adult—were converted into numerical form,adjusted as necessary,and subsequently clustered.Clustering was performed using six different methods:Kmeans,agglomerative,DBSCAN(Density-Based Spatial Clustering of Applications with Noise),mean shift,spectral,and Birch.In the second phase,the clustered ASD data were classified.The model’s accuracy was assessed using 5-fold cross-validation to ensure robust evaluation.In total,ten distinct machine learning algorithms were employed.The findings indicate that all clustering methods demonstrated success with various classifiers.Notably,the K-means algorithm emerged as particularly effective,achieving consistent and significant results across all datasets.This study is expected to serve as a guide for improving ASD detection performance,even with minimal data availability.展开更多
The continuous mutation and rapid spread of the severe acute respiratory syndrome coronavirus 2(SARSCoV-2)have led to the ineffectiveness of many antiviral drugs targeting the original strain.To keep pace with the vi...The continuous mutation and rapid spread of the severe acute respiratory syndrome coronavirus 2(SARSCoV-2)have led to the ineffectiveness of many antiviral drugs targeting the original strain.To keep pace with the virus’evolutionary speed,there is a crucial need for the development of rapid,cost-effective,and efficient inhibitor screening methods.In this study,we created a novel approach based on fluorescence resonance energy transfer(FRET)technology for in vitro detection of inhibitors targeting the interaction between the SARS-CoV-2 spike protein RBD(s-RBD)and the virus receptor angiotensin-converting enzyme 2(ACE2).Utilizing crystallographic insights into the s-RBD/ACE2 interaction,we modified ACE2 by fusing SNAP tag to its N-terminus(resulting in SA740)and Halo tag to s-RBD’s C-terminus(producing R525H and R541H),thereby ensuring the proximity(<10 nm)of labeled FRET dyes.We found that relative to the R541H fusion protein,R525H exhibited higher FRET efficiency,which attributed to the shortened distance between FRET dyes due to the truncation of s-RBD.Utilizing the sensitive FRET effect between SA740 and R525H,we evaluated its efficacy in detecting inhibitors of SARS-CoV-2 entry in solution and live cells.Ultimately,this FRET-based detection method was demonstrated high sensitivity,rapidity,and simplicity in solution and held promise for high-throughput screening of SARS-CoV-2 inhibitors.展开更多
Nanochannel technology based on ionic current rectification has emerged as a powerful tool for the detection of biomolecules owing to unique advantages.Nevertheless,existing nanochannel sensors mainly focus on the det...Nanochannel technology based on ionic current rectification has emerged as a powerful tool for the detection of biomolecules owing to unique advantages.Nevertheless,existing nanochannel sensors mainly focus on the detection of targets in solution or inside the cells,moreover,they only have a single function,greatly limiting their application.Herein,we fabricated SuperDNA self-assembled conical nanochannel,which was clamped in the middle of self-made device for two functions:Online detecting living cells released TNF-αand studying intercellular communication.Polyethylene terephthalate(PET)membrane incubated tumor associated macrophages and tumor cells was rolled up and inserted into the left and right chamber of the device,respectively.Through monitoring the ion current change in the nanochannel,tumor associated macrophages released TNF-αcould be in situ and noninvasive detected with a detection limit of 0.23 pg/mL.Furthermore,the secreted TNF-αinduced epithelial-mesenchymal transformation of tumor cells in the right chamber was also studied.The presented strategy displayed outstanding performance and multi-function,providing a promising platform for in situ non-destructive detection of cell secretions and related intercellular communication analysis.展开更多
INTRODUCTION.On May 1st,2024,around 2:10 a.m.,a catastrophic collapse occurred along the Meilong Expressway near Meizhou City,Guangdong Province,China,at coordinates 24°29′24″N and 116°40′25″E.This colla...INTRODUCTION.On May 1st,2024,around 2:10 a.m.,a catastrophic collapse occurred along the Meilong Expressway near Meizhou City,Guangdong Province,China,at coordinates 24°29′24″N and 116°40′25″E.This collapse resulted in a pavement failure of approximately 17.9 m in length and covering an area of about 184.3 m^(2)(Chinanews,2024).展开更多
With the widespread use of blockchain technology for smart contracts and decentralized applications on the Ethereum platform, the blockchain has become a cornerstone of trust in the modern financial system. However, i...With the widespread use of blockchain technology for smart contracts and decentralized applications on the Ethereum platform, the blockchain has become a cornerstone of trust in the modern financial system. However, its anonymity has provided new ways for Ponzi schemes to commit fraud, posing significant risks to investors. Current research still has some limitations, for example, Ponzi schemes are difficult to detect in the early stages of smart contract deployment, and data imbalance is not considered. In addition, there is room for improving the detection accuracy. To address the above issues, this paper proposes LT-SPSD (LSTM-Transformer smart Ponzi schemes detection), which is a Ponzi scheme detection method that combines Long Short-Term Memory (LSTM) and Transformer considering the time-series transaction information of smart contracts as well as the global information. Based on the verified smart contract addresses, account features, and code features are extracted to construct a feature dataset, and the SMOTE-Tomek algorithm is used to deal with the imbalanced data classification problem. By comparing our method with the other four typical detection methods in the experiment, the LT-SPSD method shows significant performance improvement in precision, recall, and F1-score. The results of the experiment confirm the efficacy of the model, which has some application value in Ethereum Ponzi scheme smart contract detection.展开更多
Anomaly detection is an important research area in a diverse range of real-world applications.Although many algorithms have been proposed to address anomaly detection for numerical datasets,categorical and mixed datas...Anomaly detection is an important research area in a diverse range of real-world applications.Although many algorithms have been proposed to address anomaly detection for numerical datasets,categorical and mixed datasets remain a significant challenge,primarily because a natural distance metric is lacking.Consequently,the methods proposed in the literature implement entirely different assumptions regarding the definition of cate-gorical anomalies.This paper presents a novel categorical anomaly detection approach,offering two key con-tributions to existing methods.First,a novel surprisal-based anomaly score is introduced,which provides a more accurate assessment of anomalies by considering the full distribution of categorical values.Second,the proposed method considers complex correlations in the data beyond the pairwise interactions of features.This study proposed and tested the novel categorical surprisal anomaly detection algorithm(CSAD)by comparing and evaluating it against six competitors.The experimental results indicate that CSAD produced the best overall performance,achieving the highest average ROC-AUC and PR-AUC values of 0.8 and 0.443,respectively.Furthermore,CSAD's execution time is satisfactory even when processing large,high-dimensional datasets.展开更多
Small-drone technology has opened a range of new applications for aerial transportation. These drones leverage the Internet of Things (IoT) to offer cross-location services for navigation. However, they are susceptibl...Small-drone technology has opened a range of new applications for aerial transportation. These drones leverage the Internet of Things (IoT) to offer cross-location services for navigation. However, they are susceptible to security and privacy threats due to hardware and architectural issues. Although small drones hold promise for expansion in both civil and defense sectors, they have safety, security, and privacy threats. Addressing these challenges is crucial to maintaining the security and uninterrupted operations of these drones. In this regard, this study investigates security, and preservation concerning both the drones and Internet of Drones (IoD), emphasizing the significance of creating drone networks that are secure and can robustly withstand interceptions and intrusions. The proposed framework incorporates a weighted voting ensemble model comprising three convolutional neural network (CNN) models to enhance intrusion detection within the network. The employed CNNs are customized 1D models optimized to obtain better performance. The output from these CNNs is voted using a weighted criterion using a 0.4, 0.3, and 0.3 ratio for three CNNs, respectively. Experiments involve using multiple benchmark datasets, achieving an impressive accuracy of up to 99.89% on drone data. The proposed model shows promising results concerning precision, recall, and F1 as indicated by their obtained values of 99.92%, 99.98%, and 99.97%, respectively. Furthermore, cross-validation and performance comparison with existing works is also carried out. Findings indicate that the proposed approach offers a prospective solution for detecting security threats for aerial systems and satellite systems with high accuracy.展开更多
Detection and treatment of colorectal cancer(CRC)at an early stage is vital for long-term survival.Liquid biopsy has emerged as a promising new avenue for non-invasive screening of CRC as well as prognostication and s...Detection and treatment of colorectal cancer(CRC)at an early stage is vital for long-term survival.Liquid biopsy has emerged as a promising new avenue for non-invasive screening of CRC as well as prognostication and surveillance of minimal residual disease.Cell free DNA(cfDNA)is a promising liquid biopsy analyte and has been approved for use in clinical practice.Here,we explore the current challenges of utilizing cfDNA in the screening and prognostication of CRC but also for detecting driver mutations in healthy,presymptomatic patients with normal colonic crypts.CfDNA for the detection of cancerous or premalignant colonic lesions has already been extensively explored,however few have considered utilizing cfDNA in the detection of driver mutations in healthy patients.Theoretically,this would allow us to detect patients who are at a higher risk of tumorigenesis decades in advance of established malignancy and stratify them into higher risk groups for early-intervention screening programs.We also explore the solutions necessary to overcome the challenges that prevent liquid biopsy from entering mainstream clinical use.The potential for liquid biopsy is immense if these challenges are successfully circumvented,and can dramatically reduce CRC rates as well as improve survival in patients.展开更多
In this study,we propose Space-to-Depth and You Only Look Once Version 7(SPD-YOLOv7),an accurate and efficient method for detecting pests inmaize crops,addressing challenges such as small pest sizes,blurred images,low...In this study,we propose Space-to-Depth and You Only Look Once Version 7(SPD-YOLOv7),an accurate and efficient method for detecting pests inmaize crops,addressing challenges such as small pest sizes,blurred images,low resolution,and significant species variation across different growth stages.To improve the model’s ability to generalize and its robustness,we incorporate target background analysis,data augmentation,and processing techniques like Gaussian noise and brightness adjustment.In target detection,increasing the depth of the neural network can lead to the loss of small target information.To overcome this,we introduce the Space-to-Depth Convolution(SPD-Conv)module into the SPD-YOLOv7 framework,replacing certain convolutional layers in the traditional system backbone and head network.This modification helps retain small target features and location information.Additionally,the Efficient Layer Aggregation Network-Wide(ELAN-W)module is combined with the Convolutional Block Attention Module(CBAM)attention mechanism to extract more efficient features.Experimental results show that the enhanced YOLOv7 model achieves an accuracy of 98.38%,with an average accuracy of 99.4%,outperforming the original YOLOv7 model.These improvements represent an increase of 2.46%in accuracy and 3.19%in average accuracy.The results indicate that the enhanced YOLOv7 model is more efficient and real-time,offering valuable insights for maize pest control.展开更多
The increasing fluency of advanced language models,such as GPT-3.5,GPT-4,and the recently introduced DeepSeek,challenges the ability to distinguish between human-authored and AI-generated academic writing.This situati...The increasing fluency of advanced language models,such as GPT-3.5,GPT-4,and the recently introduced DeepSeek,challenges the ability to distinguish between human-authored and AI-generated academic writing.This situation is raising significant concerns regarding the integrity and authenticity of academic work.In light of the above,the current research evaluates the effectiveness of Bidirectional Long Short-TermMemory(BiLSTM)networks enhanced with pre-trained GloVe(Global Vectors for Word Representation)embeddings to detect AIgenerated scientific Abstracts drawn from the AI-GA(Artificial Intelligence Generated Abstracts)dataset.Two core BiLSTM variants were assessed:a single-layer approach and a dual-layer design,each tested under static or adaptive embeddings.The single-layer model achieved nearly 97%accuracy with trainable GloVe,occasionally surpassing the deeper model.Despite these gains,neither configuration fully matched the 98.7%benchmark set by an earlier LSTMWord2Vec pipeline.Some runs were over-fitted when embeddings were fine-tuned,whereas static embeddings offered a slightly lower yet stable accuracy of around 96%.This lingering gap reinforces a key ethical and procedural concern:relying solely on automated tools,such as Turnitin’s AI-detection features,to penalize individuals’risks and unjust outcomes.Misclassifications,whether legitimate work is misread as AI-generated or engineered text,evade detection,demonstrating that these classifiers should not stand as the sole arbiters of authenticity.Amore comprehensive approach is warranted,one which weaves model outputs into a systematic process supported by expert judgment and institutional guidelines designed to protect originality.展开更多
Listeria monocytogenes(LM)is a dangerous foodborne pathogen for humans.One emerging and validated method of indirectly assessing LM in food is detecting 3-hydroxy-2-butanone(3H2B)gas.In this study,the synthesis of 3-(...Listeria monocytogenes(LM)is a dangerous foodborne pathogen for humans.One emerging and validated method of indirectly assessing LM in food is detecting 3-hydroxy-2-butanone(3H2B)gas.In this study,the synthesis of 3-(2-aminoethylamino)propyltrimethoxysilane(AAPTMS)functionalized hierarchical hollow TiO_(2)nanospheres was achieved via precise controlling of solvothermal reaction temperature and post-grafting route.The sensors based on as-prepared materials exhibited excellent sensitivity(480 Hz@50 ppm),low detection limit(100 ppb),and outstanding selectivity.Moreover,the evaluation of LM with high sensitivity and specificity was achieved using the sensors.Such stable three-dimensional spheres,whose distinctive hierarchical and hollow nanostructure simultaneously improved both sensitivity and response/recovery speed dramatically,were spontaneously assembled by nanosheets.Meanwhile,the moderate loadings of AAPTMS significantly improved the selectivity of sensors.Then,the gas-sensing mechanism was explored by utilizing thermodynamic investigation,Gaussian 16 software,and in situ diffuse reflectance infrared transform spectroscopy,illustrating the weak chemisorption between the-NHgroup and 3H2B molecules.These portable sensors are promising for real-time assessment of LM at room temperature,which will make a magnificent contribution to food safety.展开更多
Atmospheric turbulence is an important parameter affecting laser atmospheric transmission.This paper reports on a self-developed atmospheric turbulence detection Li DAR system(scanning differential image motion Li DAR...Atmospheric turbulence is an important parameter affecting laser atmospheric transmission.This paper reports on a self-developed atmospheric turbulence detection Li DAR system(scanning differential image motion Li DAR(DIM-Li DAR)system).By designing and simulating the optical system of atmospheric turbulence detection Li DAR,the basic optical imaging accuracy has been determined.展开更多
The potential of text analytics is revealed by Machine Learning(ML)and Natural Language Processing(NLP)techniques.In this paper,we propose an NLP framework that is applied to multiple datasets to detect malicious Unif...The potential of text analytics is revealed by Machine Learning(ML)and Natural Language Processing(NLP)techniques.In this paper,we propose an NLP framework that is applied to multiple datasets to detect malicious Uniform Resource Locators(URLs).Three categories of features,both ML and Deep Learning(DL)algorithms and a ranking schema are included in the proposed framework.We apply frequency and prediction-based embeddings,such as hash vectorizer,Term Frequency-Inverse Dense Frequency(TF-IDF)and predictors,word to vector-word2vec(continuous bag of words,skip-gram)from Google,to extract features from text.Further,we apply more state-of-the-art methods to create vectorized features,such as GloVe.Additionally,feature engineering that is specific to URL structure is deployed to detect scams and other threats.For framework assessment,four ranking indicators are weighted:computational time and performance as accuracy,F1 score and type error II.For the computational time,we propose a new metric-Feature Building Time(FBT)as the cutting-edge feature builders(like doc2vec or GloVe)require more time.By applying the proposed assessment step,the skip-gram algorithm of word2vec surpasses other feature builders in performance.Additionally,eXtreme Gradient Boost(XGB)outperforms other classifiers.With this setup,we attain an accuracy of 99.5%and an F1 score of 0.99.展开更多
Wearing helmetswhile riding electric bicycles can significantly reduce head injuries resulting fromtraffic accidents.To effectively monitor compliance,the utilization of target detection algorithms through traffic cam...Wearing helmetswhile riding electric bicycles can significantly reduce head injuries resulting fromtraffic accidents.To effectively monitor compliance,the utilization of target detection algorithms through traffic cameras plays a vital role in identifying helmet usage by electric bicycle riders and recognizing license plates on electric bicycles.However,manual enforcement by traffic police is time-consuming and labor-intensive.Traditional methods face challenges in accurately identifying small targets such as helmets and license plates using deep learning techniques.This paper proposes an enhanced model for detecting helmets and license plates on electric bicycles,addressing these challenges.The proposedmodel improves uponYOLOv8n by deepening the network structure,incorporating weighted connections,and introducing lightweight convolutional modules.These modifications aim to enhance the precision of small target recognition while reducing the model’s parameters,making it suitable for deployment on low-performance devices in real traffic scenarios.Experimental results demonstrate that the model achieves an mAP@0.5 of 91.8%,showing an 11.5%improvement over the baselinemodel,with a 16.2%reduction in parameters.Additionally,themodel achieves a frames per second(FPS)rate of 58,meeting the accuracy and speed requirements for detection in actual traffic scenarios.展开更多
文摘At an age when most teens are figuring out high school,Siddharth is already shaping the future of medical tech.The 14⁃year⁃old boy from Dallas has created an AI⁃powered app,Circadian AI,capable of detecting heart disease in just 7 seconds using only a smartphone's microphone.
文摘BACKGROUND Diagnosing bacterial infections(BI)in patients with cirrhosis can be challenging because of unclear symptoms,low diagnostic accuracy,and lengthy culture testing times.Various biomarkers have been studied,including serum procal-citonin(PCT)and presepsin.However,the diagnostic performance of these markers remains unclear,requiring further informative studies to ascertain their diagnostic value.AIM To evaluate the pooled diagnostic performance of PCT and presepsin in detecting BI among patients with cirrhosis.INTRODUCTION Bacterial infections(BI)commonly occur in patients with cirrhosis,resulting in poor outcomes,including the development of cirrhotic complications,septic shock,acute-on-chronic liver failure(ACLF),multiple organ failures,and mortality[1,2].BI is observed in 20%-30%of hospitalized patients,with and without ACLF[3].Patients with cirrhosis are susceptible to BI because of internal and external factors.The major internal factors are changes in gut microbial composition and function,bacterial translocation,and cirrhosis-associated immune dysfunction syndrome[4,5].External factors include alcohol use,proton-pump inhibitor use,frailty,readmission,and invasive procedures.Spontaneous bacterial peritonitis(SBP),urinary tract infection,pneumonia,and primary bacteremia are the common BIs in hospit-alized patients with cirrhosis[6].Early diagnosis and adequate empirical antibiotic therapy are two critical factors that improve the prognosis of BI in patients with cirrhosis.However,early detection of BI in cirrhosis is challenging due to subtle clinical signs and symptoms,low sensitivity and specificity of systemic inflammatory response syndrome criteria,and low sensitivity of bacterial cultures.Thus,effective biomarkers need to be identified for the early detection of BI.Several biomarkers have been evaluated,but their efficacy in detecting BI is unclear.Procalcitonin(PCT)is a precursor of the hormone calcitonin,which is secreted by parafollicular cells of the thyroid gland[7].In the presence of BI,PCT gene expression increases in extrathyroidal tissues,causing a subsequent increase in serum PCT level[8].Changes in serum PCT are detectable as early as 4 hours after infection onset and peaks between 8 and 24 hours,making it a valuable diagnostic biomarker for BI.Several studies have demonstrated the favorable diagnostic accuracy of PCT in the diagnosis of BI in individuals with cirrhosis[9-13]and without cirrhosis[14-16].Since 2014,two meta-analyses have been published on the diagnostic value of PCT for SBP and BI in patients with cirrhosis[17,18].Other related studies have been conducted since then[10-12,19-33].Serum presepsin has recently emerged as a promising biomarker for diagnosing BI.This biomarker is the N-terminal fraction protein of the soluble CD14 g-negative bacterial lipopolysaccharide–lipopolysaccharide binding protein(sCD14-LPS-LBP)complex,which is cleaved by inflammatory serum protease in response to BI[34].Presepsin levels increase within 2 hours and peaks in 3 hours[35].This is useful for detecting BI since presepsin levels increase earlier than serum Our systematic review and meta-analysis was performed with adherence to PRISMA guidelines[37].
文摘Tomato plant diseases often first manifest on the leaves,making the detection of tomato leaf diseases particularly crucial for the tomato cultivation industry.However,conventional deep learning models face challenges such as large model sizes and slow detection speeds when deployed on resource-constrained platforms and agricultural machinery.This paper proposes a lightweight model for detecting tomato leaf diseases,named LT-YOLO,based on the YOLOv8n architecture.First,we enhance the C2f module into a RepViT Block(RVB)with decoupled token and channel mixers to reduce the cost of feature extraction.Next,we incorporate a novel Efficient Multi-Scale Attention(EMA)mechanism in the deeper layers of the backbone to improve detection of critical disease features.Additionally,we design a lightweight detection head,LT-Detect,using Partial Convolution(PConv)to significantly reduce the classification and localization costs during detection.Finally,we introduce a Receptive Field Block(RFB)in the shallow layers of the backbone to expand the model’s receptive field,enabling effective detection of diseases at various scales.The improved model reduces the number of parameters by 43%and the computational load by 50%.Additionally,it achieves a mean Average Precision(mAP)of 90.9%on a publicly available dataset containing 3641 images of tomato leaf diseases,with only a 0.7%decrease compared to the baseline model.This demonstrates that the model maintains excellent accuracy while being lightweight,making it suitable for rapid detection of tomato leaf diseases.
文摘Driver distraction is a leading cause of traffic accidents,with fatigue being a significant contributor.This paper introduces a novel method for detecting driver distraction by analyzing facial features using machine deep learning and 68 face model.The proposed system assesses driver tiredness by measuring the distance between key facial landmarks,such as the distance between the eyes and the angle of the mouth,to evaluate signs of drowsiness or disengagement.Real-time video feed analysis allows for continuous monitoring of the driver’s face,enabling the system to detect behavioral cues associated with distraction,such as eye closures or changes in facial expressions.The effectiveness of this method is demonstrated through a series of experiments on a dataset of driver videos,which proves that the approach can accurately assess tiredness and distraction levels under various driving conditions.By focusing on facial landmarks,the system is computationally efficient and capable of operating in real-time,making it a practical solution for in-vehicle safety systems.This paper discusses the system’s performance,limitations,and potential for future enhancements,including integration with other in-vehicle technologies to provide comprehensive driver monitoring.
基金supported by the Department of science and technology of Shaanxi Province(NO.2023-ZDLGY-24).
文摘Detection of ore waste is crucial for achieving automation in mineral metallurgy production.However,deep learning-based target detection algorithms still face several challenges in iron waste screening,including poor lighting conditions in underground mining environments,dust disturbances,platform vibrations during operation,and limited resources for large-scale computing equipment.These factors contribute to extended computation times and unsatisfactory detection accuracy.To address these challenges,this paper proposes an ore waste detection algorithm based on an improved version of YOLOv5.To enhance feature extraction capabilities,the RepLKNet module is incorporated into the YOLOv5 backbone and neck networks.This module enhances the deformation information of feature extraction with the maximum effective Receptive Field to increase the model's accuracy.The Normalizationbased Attention Module(NAM)was introduced to enhance the attention mechanism by focusing on the most relevant features.This improves accuracy in detecting objects against noisy or unclear backgrounds,thereby further enhancing detection performance while reducing model parameters.Additionally,the loss function is optimized to constrain angular deviation using the SIOU loss function,which prevents the training frame from drifting during training and enhances convergence speed.To validate the performance of the proposed method,we tested it using a self-constructed dataset comprising 1,328 images obtained from the crushing station at Jinchuan Group's No.2 mine.The results indicate that,compared to YOLOv5s on the self-constructed dataset,the proposed algorithm achieves an 18.3%improvement in mAP(0.5),a 54%reduction in FLOPs,and a 52.53%decrease in model parameters.The effectiveness and superiority of the proposed algorithm are demonstrated through case studies and comparative analyses.
基金supported by the STI 2030 Major Projects(2022ZD0211700)the Key R&D Program of Guangdong Province,China(2018B030339001)+2 种基金the Key Realm R&D Program of Guangzhou,China(202007030007)the National Natural Science Foundation of China(82371538)The authors gratefully acknowledge the approval granted by the Ethics Committee of the Affiliated Brain Hospital of Guangzhou Medical University for this study involving human participants,with the approval ID(2021)No.071.
文摘Depression is increasingly prevalent among adolescents and can profoundly impact their lives.However,the early detection of depression is often hindered by the timeconsuming diagnostic process and the absence of objective biomarkers.In this study,we propose a novel approach for depression detection based on an affective brain-computer interface(aBCI)and the resting-state electroencephalogram(EEG).By fusing EEG features associated with both emotional and resting states,our method captures comprehensive depression-related information.The final depression detection model,derived through decision fusion with multiple independent models,further enhances detection efficacy.Our experiments involved 40 adolescents with depression and 40 matched controls.The proposed model achieved an accuracy of 86.54%on cross-validation and 88.20%on the independent test set,demonstrating the efficiency of multi-modal fusion.In addition,further analysis revealed distinct brain activity patterns between the two groups across different modalities.These findings hold promise for new directions in depression detection and intervention.
文摘Detection of Autism Spectrum Disorder(ASD)is a crucial area of research,representing a foundational aspect of psychological studies.The advancement of technology and the widespread adoption of machine learning methodologies have brought significant attention to this field in recent years.Interdisciplinary efforts have further propelled research into detection methods.Consequently,this study aims to contribute to both the fields of psychology and computer science.Specifically,the goal is to apply machine learning techniques to limited data for the detection of Autism Spectrum Disorder.This study is structured into two distinct phases:data preprocessing and classification.In the data preprocessing phase,four datasets—Toddler,Children,Adolescent,and Adult—were converted into numerical form,adjusted as necessary,and subsequently clustered.Clustering was performed using six different methods:Kmeans,agglomerative,DBSCAN(Density-Based Spatial Clustering of Applications with Noise),mean shift,spectral,and Birch.In the second phase,the clustered ASD data were classified.The model’s accuracy was assessed using 5-fold cross-validation to ensure robust evaluation.In total,ten distinct machine learning algorithms were employed.The findings indicate that all clustering methods demonstrated success with various classifiers.Notably,the K-means algorithm emerged as particularly effective,achieving consistent and significant results across all datasets.This study is expected to serve as a guide for improving ASD detection performance,even with minimal data availability.
基金supported by the National Natural Science Foundation of China(Nos.22225806,22378385,22078314,22278394)Dalian Institute of Chemical Physics(Nos.DICPI202142,DICPI202436,DMU-1&DICP UN202301).
文摘The continuous mutation and rapid spread of the severe acute respiratory syndrome coronavirus 2(SARSCoV-2)have led to the ineffectiveness of many antiviral drugs targeting the original strain.To keep pace with the virus’evolutionary speed,there is a crucial need for the development of rapid,cost-effective,and efficient inhibitor screening methods.In this study,we created a novel approach based on fluorescence resonance energy transfer(FRET)technology for in vitro detection of inhibitors targeting the interaction between the SARS-CoV-2 spike protein RBD(s-RBD)and the virus receptor angiotensin-converting enzyme 2(ACE2).Utilizing crystallographic insights into the s-RBD/ACE2 interaction,we modified ACE2 by fusing SNAP tag to its N-terminus(resulting in SA740)and Halo tag to s-RBD’s C-terminus(producing R525H and R541H),thereby ensuring the proximity(<10 nm)of labeled FRET dyes.We found that relative to the R541H fusion protein,R525H exhibited higher FRET efficiency,which attributed to the shortened distance between FRET dyes due to the truncation of s-RBD.Utilizing the sensitive FRET effect between SA740 and R525H,we evaluated its efficacy in detecting inhibitors of SARS-CoV-2 entry in solution and live cells.Ultimately,this FRET-based detection method was demonstrated high sensitivity,rapidity,and simplicity in solution and held promise for high-throughput screening of SARS-CoV-2 inhibitors.
基金supported by National Natural Science Foundation of China(Nos.22174016,22374019,and 22209025)Natural Science Foundation of Jiangsu Province(No.BK20220799).
文摘Nanochannel technology based on ionic current rectification has emerged as a powerful tool for the detection of biomolecules owing to unique advantages.Nevertheless,existing nanochannel sensors mainly focus on the detection of targets in solution or inside the cells,moreover,they only have a single function,greatly limiting their application.Herein,we fabricated SuperDNA self-assembled conical nanochannel,which was clamped in the middle of self-made device for two functions:Online detecting living cells released TNF-αand studying intercellular communication.Polyethylene terephthalate(PET)membrane incubated tumor associated macrophages and tumor cells was rolled up and inserted into the left and right chamber of the device,respectively.Through monitoring the ion current change in the nanochannel,tumor associated macrophages released TNF-αcould be in situ and noninvasive detected with a detection limit of 0.23 pg/mL.Furthermore,the secreted TNF-αinduced epithelial-mesenchymal transformation of tumor cells in the right chamber was also studied.The presented strategy displayed outstanding performance and multi-function,providing a promising platform for in situ non-destructive detection of cell secretions and related intercellular communication analysis.
基金supported by the National Natural Science Foundation of China(Nos.42371094,41907253)partially supported by the Interdisciplinary Cultivation Program of Xidian University(No.21103240005)the Postdoctoral Fellowship Program of CPSF(No.GZB20240589)。
文摘INTRODUCTION.On May 1st,2024,around 2:10 a.m.,a catastrophic collapse occurred along the Meilong Expressway near Meizhou City,Guangdong Province,China,at coordinates 24°29′24″N and 116°40′25″E.This collapse resulted in a pavement failure of approximately 17.9 m in length and covering an area of about 184.3 m^(2)(Chinanews,2024).
基金This work was granted by Qin Xin Talents Cultivation Program(No.QXTCP C202115)Beijing Information Science and Technology University+1 种基金the Beijing Advanced Innovation Center for Future Blockchain and Privacy Computing Fund(No.GJJ-23)National Social Science Foundation,China(No.21BTQ079).
文摘With the widespread use of blockchain technology for smart contracts and decentralized applications on the Ethereum platform, the blockchain has become a cornerstone of trust in the modern financial system. However, its anonymity has provided new ways for Ponzi schemes to commit fraud, posing significant risks to investors. Current research still has some limitations, for example, Ponzi schemes are difficult to detect in the early stages of smart contract deployment, and data imbalance is not considered. In addition, there is room for improving the detection accuracy. To address the above issues, this paper proposes LT-SPSD (LSTM-Transformer smart Ponzi schemes detection), which is a Ponzi scheme detection method that combines Long Short-Term Memory (LSTM) and Transformer considering the time-series transaction information of smart contracts as well as the global information. Based on the verified smart contract addresses, account features, and code features are extracted to construct a feature dataset, and the SMOTE-Tomek algorithm is used to deal with the imbalanced data classification problem. By comparing our method with the other four typical detection methods in the experiment, the LT-SPSD method shows significant performance improvement in precision, recall, and F1-score. The results of the experiment confirm the efficacy of the model, which has some application value in Ethereum Ponzi scheme smart contract detection.
文摘Anomaly detection is an important research area in a diverse range of real-world applications.Although many algorithms have been proposed to address anomaly detection for numerical datasets,categorical and mixed datasets remain a significant challenge,primarily because a natural distance metric is lacking.Consequently,the methods proposed in the literature implement entirely different assumptions regarding the definition of cate-gorical anomalies.This paper presents a novel categorical anomaly detection approach,offering two key con-tributions to existing methods.First,a novel surprisal-based anomaly score is introduced,which provides a more accurate assessment of anomalies by considering the full distribution of categorical values.Second,the proposed method considers complex correlations in the data beyond the pairwise interactions of features.This study proposed and tested the novel categorical surprisal anomaly detection algorithm(CSAD)by comparing and evaluating it against six competitors.The experimental results indicate that CSAD produced the best overall performance,achieving the highest average ROC-AUC and PR-AUC values of 0.8 and 0.443,respectively.Furthermore,CSAD's execution time is satisfactory even when processing large,high-dimensional datasets.
文摘Small-drone technology has opened a range of new applications for aerial transportation. These drones leverage the Internet of Things (IoT) to offer cross-location services for navigation. However, they are susceptible to security and privacy threats due to hardware and architectural issues. Although small drones hold promise for expansion in both civil and defense sectors, they have safety, security, and privacy threats. Addressing these challenges is crucial to maintaining the security and uninterrupted operations of these drones. In this regard, this study investigates security, and preservation concerning both the drones and Internet of Drones (IoD), emphasizing the significance of creating drone networks that are secure and can robustly withstand interceptions and intrusions. The proposed framework incorporates a weighted voting ensemble model comprising three convolutional neural network (CNN) models to enhance intrusion detection within the network. The employed CNNs are customized 1D models optimized to obtain better performance. The output from these CNNs is voted using a weighted criterion using a 0.4, 0.3, and 0.3 ratio for three CNNs, respectively. Experiments involve using multiple benchmark datasets, achieving an impressive accuracy of up to 99.89% on drone data. The proposed model shows promising results concerning precision, recall, and F1 as indicated by their obtained values of 99.92%, 99.98%, and 99.97%, respectively. Furthermore, cross-validation and performance comparison with existing works is also carried out. Findings indicate that the proposed approach offers a prospective solution for detecting security threats for aerial systems and satellite systems with high accuracy.
文摘Detection and treatment of colorectal cancer(CRC)at an early stage is vital for long-term survival.Liquid biopsy has emerged as a promising new avenue for non-invasive screening of CRC as well as prognostication and surveillance of minimal residual disease.Cell free DNA(cfDNA)is a promising liquid biopsy analyte and has been approved for use in clinical practice.Here,we explore the current challenges of utilizing cfDNA in the screening and prognostication of CRC but also for detecting driver mutations in healthy,presymptomatic patients with normal colonic crypts.CfDNA for the detection of cancerous or premalignant colonic lesions has already been extensively explored,however few have considered utilizing cfDNA in the detection of driver mutations in healthy patients.Theoretically,this would allow us to detect patients who are at a higher risk of tumorigenesis decades in advance of established malignancy and stratify them into higher risk groups for early-intervention screening programs.We also explore the solutions necessary to overcome the challenges that prevent liquid biopsy from entering mainstream clinical use.The potential for liquid biopsy is immense if these challenges are successfully circumvented,and can dramatically reduce CRC rates as well as improve survival in patients.
文摘In this study,we propose Space-to-Depth and You Only Look Once Version 7(SPD-YOLOv7),an accurate and efficient method for detecting pests inmaize crops,addressing challenges such as small pest sizes,blurred images,low resolution,and significant species variation across different growth stages.To improve the model’s ability to generalize and its robustness,we incorporate target background analysis,data augmentation,and processing techniques like Gaussian noise and brightness adjustment.In target detection,increasing the depth of the neural network can lead to the loss of small target information.To overcome this,we introduce the Space-to-Depth Convolution(SPD-Conv)module into the SPD-YOLOv7 framework,replacing certain convolutional layers in the traditional system backbone and head network.This modification helps retain small target features and location information.Additionally,the Efficient Layer Aggregation Network-Wide(ELAN-W)module is combined with the Convolutional Block Attention Module(CBAM)attention mechanism to extract more efficient features.Experimental results show that the enhanced YOLOv7 model achieves an accuracy of 98.38%,with an average accuracy of 99.4%,outperforming the original YOLOv7 model.These improvements represent an increase of 2.46%in accuracy and 3.19%in average accuracy.The results indicate that the enhanced YOLOv7 model is more efficient and real-time,offering valuable insights for maize pest control.
文摘The increasing fluency of advanced language models,such as GPT-3.5,GPT-4,and the recently introduced DeepSeek,challenges the ability to distinguish between human-authored and AI-generated academic writing.This situation is raising significant concerns regarding the integrity and authenticity of academic work.In light of the above,the current research evaluates the effectiveness of Bidirectional Long Short-TermMemory(BiLSTM)networks enhanced with pre-trained GloVe(Global Vectors for Word Representation)embeddings to detect AIgenerated scientific Abstracts drawn from the AI-GA(Artificial Intelligence Generated Abstracts)dataset.Two core BiLSTM variants were assessed:a single-layer approach and a dual-layer design,each tested under static or adaptive embeddings.The single-layer model achieved nearly 97%accuracy with trainable GloVe,occasionally surpassing the deeper model.Despite these gains,neither configuration fully matched the 98.7%benchmark set by an earlier LSTMWord2Vec pipeline.Some runs were over-fitted when embeddings were fine-tuned,whereas static embeddings offered a slightly lower yet stable accuracy of around 96%.This lingering gap reinforces a key ethical and procedural concern:relying solely on automated tools,such as Turnitin’s AI-detection features,to penalize individuals’risks and unjust outcomes.Misclassifications,whether legitimate work is misread as AI-generated or engineered text,evade detection,demonstrating that these classifiers should not stand as the sole arbiters of authenticity.Amore comprehensive approach is warranted,one which weaves model outputs into a systematic process supported by expert judgment and institutional guidelines designed to protect originality.
基金supported by the National Natural Science Foundation of China(No.32272399)the Shanghai Natural Science Foundation(No.21ZR1427500).
文摘Listeria monocytogenes(LM)is a dangerous foodborne pathogen for humans.One emerging and validated method of indirectly assessing LM in food is detecting 3-hydroxy-2-butanone(3H2B)gas.In this study,the synthesis of 3-(2-aminoethylamino)propyltrimethoxysilane(AAPTMS)functionalized hierarchical hollow TiO_(2)nanospheres was achieved via precise controlling of solvothermal reaction temperature and post-grafting route.The sensors based on as-prepared materials exhibited excellent sensitivity(480 Hz@50 ppm),low detection limit(100 ppb),and outstanding selectivity.Moreover,the evaluation of LM with high sensitivity and specificity was achieved using the sensors.Such stable three-dimensional spheres,whose distinctive hierarchical and hollow nanostructure simultaneously improved both sensitivity and response/recovery speed dramatically,were spontaneously assembled by nanosheets.Meanwhile,the moderate loadings of AAPTMS significantly improved the selectivity of sensors.Then,the gas-sensing mechanism was explored by utilizing thermodynamic investigation,Gaussian 16 software,and in situ diffuse reflectance infrared transform spectroscopy,illustrating the weak chemisorption between the-NHgroup and 3H2B molecules.These portable sensors are promising for real-time assessment of LM at room temperature,which will make a magnificent contribution to food safety.
基金jointly funded by the National Science Foundation of China(No.42405069)the University Natural Sciences Research Project of Anhui Province(Nos.2023AH052201 and 2023AH052184)+1 种基金the 2023 Talent Research Fund Project of Hefei University(No.23RC01)the Technical Development Project of Hefei University(Nos.902/22050124128,902/22050124148 and 902/22050124250)。
文摘Atmospheric turbulence is an important parameter affecting laser atmospheric transmission.This paper reports on a self-developed atmospheric turbulence detection Li DAR system(scanning differential image motion Li DAR(DIM-Li DAR)system).By designing and simulating the optical system of atmospheric turbulence detection Li DAR,the basic optical imaging accuracy has been determined.
基金supported by a grant of the Ministry of Research,Innovation and Digitization,CNCS-UEFISCDI,Project Number PN-Ⅲ-P4-PCE-2021-0334,within PNCDI Ⅲ.
文摘The potential of text analytics is revealed by Machine Learning(ML)and Natural Language Processing(NLP)techniques.In this paper,we propose an NLP framework that is applied to multiple datasets to detect malicious Uniform Resource Locators(URLs).Three categories of features,both ML and Deep Learning(DL)algorithms and a ranking schema are included in the proposed framework.We apply frequency and prediction-based embeddings,such as hash vectorizer,Term Frequency-Inverse Dense Frequency(TF-IDF)and predictors,word to vector-word2vec(continuous bag of words,skip-gram)from Google,to extract features from text.Further,we apply more state-of-the-art methods to create vectorized features,such as GloVe.Additionally,feature engineering that is specific to URL structure is deployed to detect scams and other threats.For framework assessment,four ranking indicators are weighted:computational time and performance as accuracy,F1 score and type error II.For the computational time,we propose a new metric-Feature Building Time(FBT)as the cutting-edge feature builders(like doc2vec or GloVe)require more time.By applying the proposed assessment step,the skip-gram algorithm of word2vec surpasses other feature builders in performance.Additionally,eXtreme Gradient Boost(XGB)outperforms other classifiers.With this setup,we attain an accuracy of 99.5%and an F1 score of 0.99.
基金supported by the Ningxia Key Research and Development Program(Talent Introduction Special Project)Project(2022YCZX0013)North Minzu University 2022 School-Level Scientific Research Platform“Digital Agriculture Enabling Ningxia Rural Revitalization Innovation Team”(2022PT_S10)+1 种基金Yinchuan City University-Enterprise Joint Innovation Project(2022XQZD009)Ningxia Key Research and Development Program(Key Project)Project(2023BDE02001).
文摘Wearing helmetswhile riding electric bicycles can significantly reduce head injuries resulting fromtraffic accidents.To effectively monitor compliance,the utilization of target detection algorithms through traffic cameras plays a vital role in identifying helmet usage by electric bicycle riders and recognizing license plates on electric bicycles.However,manual enforcement by traffic police is time-consuming and labor-intensive.Traditional methods face challenges in accurately identifying small targets such as helmets and license plates using deep learning techniques.This paper proposes an enhanced model for detecting helmets and license plates on electric bicycles,addressing these challenges.The proposedmodel improves uponYOLOv8n by deepening the network structure,incorporating weighted connections,and introducing lightweight convolutional modules.These modifications aim to enhance the precision of small target recognition while reducing the model’s parameters,making it suitable for deployment on low-performance devices in real traffic scenarios.Experimental results demonstrate that the model achieves an mAP@0.5 of 91.8%,showing an 11.5%improvement over the baselinemodel,with a 16.2%reduction in parameters.Additionally,themodel achieves a frames per second(FPS)rate of 58,meeting the accuracy and speed requirements for detection in actual traffic scenarios.
