Encrypted traffic classification has become a hot issue in network security research.The class imbalance problem of traffic samples often causes the deterioration of Machine Learning based classifier performance.Altho...Encrypted traffic classification has become a hot issue in network security research.The class imbalance problem of traffic samples often causes the deterioration of Machine Learning based classifier performance.Although the Generative Adversarial Network(GAN)method can generate new samples by learning the feature distribution of the original samples,it is confronted with the problems of unstable training andmode collapse.To this end,a novel data augmenting approach called Graph CWGAN-GP is proposed in this paper.The traffic data is first converted into grayscale images as the input for the proposed model.Then,the minority class data is augmented with our proposed model,which is built by introducing conditional constraints and a new distance metric in typical GAN.Finally,the classical deep learning model is adopted as a classifier to classify datasets augmented by the Condition GAN(CGAN),Wasserstein GAN-Gradient Penalty(WGAN-GP)and Graph CWGAN-GP,respectively.Compared with the state-of-the-art GAN methods,the Graph CWGAN-GP cannot only control the modes of the data to be generated,but also overcome the problem of unstable training and generate more realistic and diverse samples.The experimental results show that the classification precision,recall and F1-Score of theminority class in the balanced dataset augmented in this paper have improved by more than 2.37%,3.39% and 4.57%,respectively.展开更多
Outcomes following peripheral nerve injury remain frustratingly poor. The reasons for this are multifactorial, although maintaining a growth permissive environment in the distal nerve stump following repair is arguabl...Outcomes following peripheral nerve injury remain frustratingly poor. The reasons for this are multifactorial, although maintaining a growth permissive environment in the distal nerve stump following repair is arguably the most important. The optimal environment for axonal regeneration relies on the synthesis and release of many biochemical mediators that are temporally and spatially regulated with a high level of incompletely understood complexity. The Schwann cell(SC) has emerged as a key player in this process. Prolonged periods of distal nerve stump denervation, characteristic of large gaps and proximal injuries, have been associated with a reduction in SC number and ability to support regenerating axons. Cell based therapy offers a potential therapy for the improvement of outcomes following peripheral nerve reconstruction. Stem cells have the potential to increase the number of SCs and prolong their ability to support regeneration. They may also have the ability to rescue and replenish populations of chromatolytic and apoptotic neurons following axotomy. Finally, they can be used in non-physiologic ways to preserve injured tissues such as denervated muscle while neuronal ingrowth has not yet occurred. Aside from stem cell type, careful consideration must be given to differentiation status, how stem cells are supported following transplantation and how they will be delivered to the site of injury. It is the aim of this article to review current opinions on the strategies of stem cell based therapy for the augmentation of peripheral nerve regeneration.展开更多
When it comes to smart healthcare business systems,network-based intrusion detection systems are crucial for protecting the system and its networks from malicious network assaults.To protect IoMT devices and networks ...When it comes to smart healthcare business systems,network-based intrusion detection systems are crucial for protecting the system and its networks from malicious network assaults.To protect IoMT devices and networks in healthcare and medical settings,our proposed model serves as a powerful tool for monitoring IoMT networks.This study presents a robust methodology for intrusion detection in Internet of Medical Things(IoMT)environments,integrating data augmentation,feature selection,and ensemble learning to effectively handle IoMT data complexity.Following rigorous preprocessing,including feature extraction,correlation removal,and Recursive Feature Elimi-nation(RFE),selected features are standardized and reshaped for deep learning models.Augmentation using the BAT algorithm enhances dataset variability.Three deep learning models,Transformer-based neural networks,self-attention Deep Convolutional Neural Networks(DCNNs),and Long Short-Term Memory(LSTM)networks,are trained to capture diverse data aspects.Their predictions form a meta-feature set for a subsequent meta-learner,which combines model strengths.Conventional classifiers validate meta-learner features for broad algorithm suitability.This comprehensive method demonstrates high accuracy and robustness in IoMT intrusion detection.Evaluations were conducted using two datasets:the publicly available WUSTL-EHMS-2020 dataset,which contains two distinct categories,and the CICIoMT2024 dataset,encompassing sixteen categories.Experimental results showcase the method’s exceptional performance,achieving optimal scores of 100%on the WUSTL-EHMS-2020 dataset and 99%on the CICIoMT2024.展开更多
Android malware has evolved in various forms such as adware that continuously exposes advertisements,banking malware designed to access users’online banking accounts,and Short Message Service(SMS)malware that uses a ...Android malware has evolved in various forms such as adware that continuously exposes advertisements,banking malware designed to access users’online banking accounts,and Short Message Service(SMS)malware that uses a Command&Control(C&C)server to send malicious SMS,intercept SMS,and steal data.By using many malicious strategies,the number of malware is steadily increasing.Increasing Android malware threats numerous users,and thus,it is necessary to detect malware quickly and accurately.Each malware has distinguishable characteristics based on its actions.Therefore,security researchers have tried to categorize malware based on their behaviors by conducting the familial analysis which can help analysists to reduce the time and cost for analyzing malware.However,those studies algorithms typically used imbalanced,well-labeled open-source dataset,and thus,it is very difficult to classify some malware families which only have a few number of malware.To overcome this challenge,previous data augmentation studies augmented data by visualizing malicious codes and used them for malware analysis.However,visualization of malware can result in misclassifications because the behavior information of the malware could be compromised.In this study,we propose an android malware familial analysis system based on a data augmentation method that preserves malware behaviors to create an effective multi-class classifier for malware family analysis.To this end,we analyze malware and use Application Programming Interface(APIs)and permissions that can reflect the behavior of malware as features.By using these features,we augment malware dataset to enable effective malware detection while preserving original malicious behaviors.Our evaluation results demonstrate that,when a model is created by using only the augmented data,a macro-F1 score of 0.65 and accuracy of 0.63%.On the other hand,when the augmented data and original malware are used together,the evaluation results show that a macro-F1 score of 0.91 and an accuracy of 0.99%.展开更多
Adversarial training has been widely considered the most effective defense against adversarial attacks.However,recent studies have demonstrated that a large discrepancy exists in the class-wise robustness of adversari...Adversarial training has been widely considered the most effective defense against adversarial attacks.However,recent studies have demonstrated that a large discrepancy exists in the class-wise robustness of adversarial training,leading to two potential issues:firstly,the overall robustness of a model is compromised due to the weakest class;and secondly,ethical concerns arising from unequal protection and biases,where certain societal demographic groups receive less robustness in defense mechanisms.Despite these issues,solutions to address the discrepancy remain largely underexplored.In this paper,we advance beyond existing methods that focus on class-level solutions.Our investigation reveals that hard examples,identified by higher cross-entropy values,can provide more fine-grained information about the discrepancy.Furthermore,we find that enhancing the diversity of hard examples can effectively reduce the robustness gap between classes.Motivated by these observations,we propose Fair Adversarial Training(FairAT)to mitigate the discrepancy of class-wise robustness.Extensive experiments on various benchmark datasets and adversarial attacks demonstrate that FairAT outperforms state-of-the-art methods in terms of both overall robustness and fairness.For a WRN-28-10 model trained on CIFAR10,FairAT improves the average and worst-class robustness by 2.13%and 4.50%,respectively.展开更多
The satellite-based augmentation system(SBAS)provides differential and integrity augmentation services for life safety fields of aviation and navigation.However,the signal structure of SBAS is public,which incurs a ri...The satellite-based augmentation system(SBAS)provides differential and integrity augmentation services for life safety fields of aviation and navigation.However,the signal structure of SBAS is public,which incurs a risk of spoofing attacks.To improve the anti-spoofing capability of the SBAS,European Union and the United States conduct research on navigation message authentication,and promote the standardization of SBAS message authentication.For the development of Beidou satellite-based augmentation system(BDSBAS),this paper proposes navigation message authentication based on the Chinese commercial cryptographic standards.Firstly,this paper expounds the architecture and principles of the SBAS message authentication,and then carries out the design of timed efficient streaming losstolerant authentication scheme(TESLA)and elliptic curve digital signature algorithm(ECDSA)authentication schemes based on Chinese commercial cryptographic standards,message arrangement and the design of over-the-air rekeying(OTAR)message.Finally,this paper conducts a theoretical analysis of the time between authentications(TBA)and maximum authentication latency(MAL)for L5 TESLA-I and L5 ECDSA-Q,and further simulates the reception time of OTAR message,TBA and MAL from the aspects of OTAR message weight and demodulation error rate.The simulation results can provide theoretical supports for the standardization of BDSBAS message authentication.展开更多
Objective expertise evaluation of individuals,as a prerequisite stage for team formation,has been a long-term desideratum in large software development companies.With the rapid advancements in machine learning methods...Objective expertise evaluation of individuals,as a prerequisite stage for team formation,has been a long-term desideratum in large software development companies.With the rapid advancements in machine learning methods,based on reliable existing data stored in project management tools’datasets,automating this evaluation process becomes a natural step forward.In this context,our approach focuses on quantifying software developer expertise by using metadata from the task-tracking systems.For this,we mathematically formalize two categories of expertise:technology-specific expertise,which denotes the skills required for a particular technology,and general expertise,which encapsulates overall knowledge in the software industry.Afterward,we automatically classify the zones of expertise associated with each task a developer has worked on using Bidirectional Encoder Representations from Transformers(BERT)-like transformers to handle the unique characteristics of project tool datasets effectively.Finally,our method evaluates the proficiency of each software specialist across already completed projects from both technology-specific and general perspectives.The method was experimentally validated,yielding promising results.展开更多
Many spore-forming Bacillus species can cause serious human diseases,because of accidental Bacillusspore infection.Thus,developing an identification strategy with both high sensitivity and specificity is greatly in de...Many spore-forming Bacillus species can cause serious human diseases,because of accidental Bacillusspore infection.Thus,developing an identification strategy with both high sensitivity and specificity is greatly in demand.In this work,we proposed a novel approach named multi-head self-attention mechanism-guided neural network Raman platform to identify living Bacillus spores within a single-cell resolution.The multi-head self-attention mechanism-guided neural network Raman platform was created by combining single-cell Raman spectroscopy,convolutional neural network(CNN),and multi-head self-attention mechanism.To address the limited size of the original spectra dataset,Gaussian noise-based spectra augmentation was employed to increase the number of single-cell Raman spectra datasets for CNN training.Owing to the assistance of both spectra augmentation and multi-head self-attention mechanism,the obtained prediction accuracy of five Bacillus spore species was further improved from 92.29±0.82%to 99.43±0.15%.To figure out the spectra differences covered by the multi-head self-attention mechanism-guided CNN,the relative classification weight from typical Raman bands was visualized via multi-head self-attention mechanism curve.In the process of spectra augmentation from 0 to 1000,the distribution of relative classification weight varied from a discrete state to a more concentrated phase.More importantly,these highlighted four Raman bands(1017,1449,1576,and 1660 cm^(-1))were assigned large weights,showing that the spectra differences in the Raman bands produced the largest contribution to prediction accuracy.It can be foreseen that,our proposed sorting platform has great potential in accurately identifying Bacillus and its related genera species at a single-cell level.展开更多
Due to the small size of the annotated corpora and the sparsity of the event trigger words, the event coreference resolver cannot capture enough event semantics, especially the trigger semantics, to identify coreferen...Due to the small size of the annotated corpora and the sparsity of the event trigger words, the event coreference resolver cannot capture enough event semantics, especially the trigger semantics, to identify coreferential event mentions. To address the above issues, this paper proposes a trigger semantics augmentation mechanism to boost event coreference resolution. First, this mechanism performs a trigger-oriented masking strategy to pre-train a BERT (Bidirectional Encoder Representations from Transformers)-based encoder (Trigger-BERT), which is fine-tuned on a large-scale unlabeled dataset Gigaword. Second, it combines the event semantic relations from the Trigger-BERT encoder with the event interactions from the soft-attention mechanism to resolve event coreference. Experimental results on both the KBP2016 and KBP2017 datasets show that our proposed model outperforms several state-of-the-art baselines.展开更多
Lung cancer continues to be a leading cause of cancer-related deaths worldwide,emphasizing the critical need for improved diagnostic techniques.Early detection of lung tumors significantly increases the chances of suc...Lung cancer continues to be a leading cause of cancer-related deaths worldwide,emphasizing the critical need for improved diagnostic techniques.Early detection of lung tumors significantly increases the chances of successful treatment and survival.However,current diagnostic methods often fail to detect tumors at an early stage or to accurately pinpoint their location within the lung tissue.Single-model deep learning technologies for lung cancer detection,while beneficial,cannot capture the full range of features present in medical imaging data,leading to incomplete or inaccurate detection.Furthermore,it may not be robust enough to handle the wide variability in medical images due to different imaging conditions,patient anatomy,and tumor characteristics.To overcome these disadvantages,dual-model or multi-model approaches can be employed.This research focuses on enhancing the detection of lung cancer by utilizing a combination of two learning models:a Convolutional Neural Network(CNN)for categorization and the You Only Look Once(YOLOv8)architecture for real-time identification and pinpointing of tumors.CNNs automatically learn to extract hierarchical features from raw image data,capturing patterns such as edges,textures,and complex structures that are crucial for identifying lung cancer.YOLOv8 incorporates multiscale feature extraction,enabling the detection of tumors of varying sizes and scales within a single image.This is particularly beneficial for identifying small or irregularly shaped tumors that may be challenging to detect.Furthermore,through the utilization of cutting-edge data augmentation methods,such as Deep Convolutional Generative Adversarial Networks(DCGAN),the suggested approach can handle the issue of limited data and boost the models’ability to learn from diverse and comprehensive datasets.The combined method not only improved accuracy and localization but also ensured efficient real-time processing,which is crucial for practical clinical applications.The CNN achieved an accuracy of 97.67%in classifying lung tissues into healthy and cancerous categories.The YOLOv8 model achieved an Intersection over Union(IoU)score of 0.85 for tumor localization,reflecting high precision in detecting and marking tumor boundaries within the images.Finally,the incorporation of synthetic images generated by DCGAN led to a 10%improvement in both the CNN classification accuracy and YOLOv8 detection performance.展开更多
Photonic platforms are gradually emerging as a promising option to encounter the ever-growing demand for artificial intelligence,among which photonic time-delay reservoir computing(TDRC)is widely anticipated.While suc...Photonic platforms are gradually emerging as a promising option to encounter the ever-growing demand for artificial intelligence,among which photonic time-delay reservoir computing(TDRC)is widely anticipated.While such a computing paradigm can only employ a single photonic device as the nonlinear node for data processing,the performance highly relies on the fading memory provided by the delay feedback loop(FL),which sets a restriction on the extensibility of physical implementation,especially for highly integrated chips.Here,we present a simplified photonic scheme for more flexible parameter configurations leveraging the designed quasi-convolution coding(QC),which completely gets rid of the dependence on FL.Unlike delay-based TDRC,encoded data in QC-based RC(QRC)enables temporal feature extraction,facilitating augmented memory capabilities.Thus,our proposed QRC is enabled to deal with time-related tasks or sequential data without the implementation of FL.Furthermore,we can implement this hardware with a low-power,easily integrable vertical-cavity surface-emitting laser for high-performance parallel processing.We illustrate the concept validation through simulation and experimental comparison of QRC and TDRC,wherein the simpler-structured QRC outperforms across various benchmark tasks.Our results may underscore an auspicious solution for the hardware implementation of deep neural networks.展开更多
Objective The study of medicine formulas is a core component of traditional Chinese medicine(TCM),yet traditional learning methods often lack interactivity and contextual understanding,making it challenging for beginn...Objective The study of medicine formulas is a core component of traditional Chinese medicine(TCM),yet traditional learning methods often lack interactivity and contextual understanding,making it challenging for beginners to grasp the intricate composition rules of formulas.To address this gap,we introduce Formula-S,a situated visualization method for TCM formula learning in augmented reality(AR)and evaluate its performance.This study aims to evaluate the effectiveness of Formula-S in enhancing TCM formula learning for beginners by comparing it with traditional text-based formula learning and web-based visualization.Methods Formula-S is an interactive AR tool designed for TCM formula learning,featuring three modes(3D,Web,and Table).The dataset included TCM formulas and herb properties extracted from authoritative references,including textbook and the SymMap database.In Formula-S,the hierarchical visualization of the formulas as herbal medicine compositions,is linked to the multidimensional herb attribute visualization and embedded in the real world,where real herb samples are presented.To evaluate its effectiveness,a controlled study(n=30)was conducted.Participants who had no formal TCM knowledge were tasked with herbal medicine identification,formula composition,and recognition.In the study,participants interacted with the AR tool through HoloLens 2.Data were collected on both task performance(accuracy and response time)and user experience,with a focus on task efficiency,accuracy,and user preference across the different learning modes.Results The situated visualization method of Formula-S had comparable accuracy to other methods but shorter response time for herbal formula learning tasks.Regarding user experience,our new approach demonstrated the highest system usability and lowest task load,effectively reducing cognitive load and allowing users to complete tasks with greater ease and efficiency.Participants reported that Formula-S enhanced their learning experience through its intuitive interface and immersive AR environment,suggesting this approach offers usability advantages for TCM education.Conclusions The situated visualization method in Formula-S offers more efficient and accurate searching capabilities compared to traditional and web-based methods.Additionally,it provides superior contextual understanding of TCM formulas,making it a promising new solution for TCM learning.展开更多
The integration of image analysis through deep learning(DL)into rock classification represents a significant leap forward in geological research.While traditional methods remain invaluable for their expertise and hist...The integration of image analysis through deep learning(DL)into rock classification represents a significant leap forward in geological research.While traditional methods remain invaluable for their expertise and historical context,DL offers a powerful complement by enhancing the speed,objectivity,and precision of the classification process.This research explores the significance of image data augmentation techniques in optimizing the performance of convolutional neural networks(CNNs)for geological image analysis,particularly in the classification of igneous,metamorphic,and sedimentary rock types from rock thin section(RTS)images.This study primarily focuses on classic image augmentation techniques and evaluates their impact on model accuracy and precision.Results demonstrate that augmentation techniques like Equalize significantly enhance the model's classification capabilities,achieving an F1-Score of 0.9869 for igneous rocks,0.9884 for metamorphic rocks,and 0.9929 for sedimentary rocks,representing improvements compared to the baseline original results.Moreover,the weighted average F1-Score across all classes and techniques is 0.9886,indicating an enhancement.Conversely,methods like Distort lead to decreased accuracy and F1-Score,with an F1-Score of 0.949 for igneous rocks,0.954 for metamorphic rocks,and 0.9416 for sedimentary rocks,exacerbating the performance compared to the baseline.The study underscores the practicality of image data augmentation in geological image classification and advocates for the adoption of DL methods in this domain for automation and improved results.The findings of this study can benefit various fields,including remote sensing,mineral exploration,and environmental monitoring,by enhancing the accuracy of geological image analysis both for scientific research and industrial applications.展开更多
Breast cancer is one of the major causes of deaths in women.However,the early diagnosis is important for screening and control the mortality rate.Thus for the diagnosis of breast cancer at the early stage,a computer-a...Breast cancer is one of the major causes of deaths in women.However,the early diagnosis is important for screening and control the mortality rate.Thus for the diagnosis of breast cancer at the early stage,a computer-aided diagnosis system is highly required.Ultrasound is an important examination technique for breast cancer diagnosis due to its low cost.Recently,many learning-based techniques have been introduced to classify breast cancer using breast ultrasound imaging dataset(BUSI)datasets;however,the manual handling is not an easy process and time consuming.The authors propose an EfficientNet-integrated ResNet deep network and XAI-based framework for accurately classifying breast cancer(malignant and benign).In the initial step,data augmentation is performed to increase the number of training samples.For this purpose,three-pixel flip mathematical equations are introduced:horizontal,vertical,and 90°.Later,two pretrained deep learning models were employed,skipped some layers,and fine-tuned.Both fine-tuned models are later trained using a deep transfer learning process and extracted features from the deeper layer.Explainable artificial intelligence-based analysed the performance of trained models.After that,a new feature selection technique is proposed based on the cuckoo search algorithm called cuckoo search controlled standard error mean.This technique selects the best features and fuses using a new parallel zeropadding maximum correlated coefficient features.In the end,the selection algorithm is applied again to the fused feature vector and classified using machine learning algorithms.The experimental process of the proposed framework is conducted on a publicly available BUSI and obtained 98.4%and 98%accuracy in two different experiments.Comparing the proposed framework is also conducted with recent techniques and shows improved accuracy.In addition,the proposed framework was executed less than the original deep learning models.展开更多
Large-scale point cloud datasets form the basis for training various deep learning networks and achieving high-quality network processing tasks.Due to the diversity and robustness constraints of the data,data augmenta...Large-scale point cloud datasets form the basis for training various deep learning networks and achieving high-quality network processing tasks.Due to the diversity and robustness constraints of the data,data augmentation(DA)methods are utilised to expand dataset diversity and scale.However,due to the complex and distinct characteristics of LiDAR point cloud data from different platforms(such as missile-borne and vehicular LiDAR data),directly applying traditional 2D visual domain DA methods to 3D data can lead to networks trained using this approach not robustly achieving the corresponding tasks.To address this issue,the present study explores DA for missile-borne LiDAR point cloud using a Monte Carlo(MC)simulation method that closely resembles practical application.Firstly,the model of multi-sensor imaging system is established,taking into account the joint errors arising from the platform itself and the relative motion during the imaging process.A distortion simulation method based on MC simulation for augmenting missile-borne LiDAR point cloud data is proposed,underpinned by an analysis of combined errors between different modal sensors,achieving high-quality augmentation of point cloud data.The effectiveness of the proposed method in addressing imaging system errors and distortion simulation is validated using the imaging scene dataset constructed in this paper.Comparative experiments between the proposed point cloud DA algorithm and the current state-of-the-art algorithms in point cloud detection and single object tracking tasks demonstrate that the proposed method can improve the network performance obtained from unaugmented datasets by over 17.3%and 17.9%,surpassing SOTA performance of current point cloud DA algorithms.展开更多
Solar cell defects exhibit significant variations and multiple types,with some defect data being difficult to acquire or having small scales,posing challenges in terms of small sample and small target in defect detect...Solar cell defects exhibit significant variations and multiple types,with some defect data being difficult to acquire or having small scales,posing challenges in terms of small sample and small target in defect detection for solar cells.In order to address this issue,this paper proposes a multi-step approach for detecting the complex defects of solar cells.First,individual cell plates are extracted from electroluminescence images for block-by-block detection.Then,StyleGAN2-Ada is utilized for generative adversarial networks data augmentation to expand the number of defect samples in small sample defects.Finally,the fake dataset is combined with real dataset,and the improved YOLOv5 model is trained on this mixed dataset.Experimental results demonstrate that the proposed method achieves a superior performance in detecting the defects with small sample and small target,with the final recall rate reaching 99.7%,an increase of 3.9% compared with the unimproved model.Additionally,the precision and mean average precision are increased by 3.4% and 3.5%,respectively.Moreover,the experiments demonstrate that the improved network training on the mixed dataset can effectively enhance the detection performance of the model.The combination of these approaches significantly improves the network’s ability to detect solar cell defects.展开更多
Phantom limb pain(PLP)is not only a physical pain experience but also poses a significant challenge to mental health and quality of life.Currently,the mechanism of PLP treatment is still unclear,and there are many met...Phantom limb pain(PLP)is not only a physical pain experience but also poses a significant challenge to mental health and quality of life.Currently,the mechanism of PLP treatment is still unclear,and there are many methods with varying effects.This article starts with the application research of extended reality technology in PLP treatment,through describing the application of its branch technologies(virtual reality,augmented reality,and mixed reality technology),to lay the foundation for subsequent research,in the hope of finding advanced and effective treatment methods,and providing a basis for future product transformation.展开更多
Data augmentation plays an important role in training deep neural model by expanding the size and diversity of the dataset.Initially,data augmentation mainly involved some simple transformations of images.Later,in ord...Data augmentation plays an important role in training deep neural model by expanding the size and diversity of the dataset.Initially,data augmentation mainly involved some simple transformations of images.Later,in order to increase the diversity and complexity of data,more advanced methods appeared and evolved to sophisticated generative models.However,these methods required a mass of computation of training or searching.In this paper,a novel training-free method that utilises the Pre-Trained Segment Anything Model(SAM)model as a data augmentation tool(PTSAM-DA)is proposed to generate the augmented annotations for images.Without the need for training,it obtains prompt boxes from the original annotations and then feeds the boxes to the pre-trained SAM to generate diverse and improved annotations.In this way,annotations are augmented more ingenious than simple manipulations without incurring huge computation for training a data augmentation model.Multiple comparative experiments on three datasets are conducted,including an in-house dataset,ADE20K and COCO2017.On this in-house dataset,namely Agricultural Plot Segmentation Dataset,maximum improvements of 3.77%and 8.92%are gained in two mainstream metrics,mIoU and mAcc,respectively.Consequently,large vision models like SAM are proven to be promising not only in image segmentation but also in data augmentation.展开更多
In order to address the widespread data shortage problem in battery research,this paper proposes a generative adversarial network model that combines it with deep convolutional networks,the Wasserstein distance,and th...In order to address the widespread data shortage problem in battery research,this paper proposes a generative adversarial network model that combines it with deep convolutional networks,the Wasserstein distance,and the gradient penalty to achieve data augmentation.To lower the threshold for implementing the proposed method,transfer learning is further introduced.The W-DC-GAN-GP-TL framework is thereby formed.This framework is evaluated on 3 different publicly available datasets to judge the quality of generated data.Through visual comparisons and the examination of two visualization methods(probability density function(PDF)and principal component analysis(PCA)),it is demonstrated that the generated data is hard to distinguish from the real data.The application of generated data for training a battery state model using transfer learning is further evaluated.Specifically,Bi-GRU-based and Transformer-based methods are implemented on 2 separate datasets for estimating state of health(SOH)and state of charge(SOC),respectively.The results indicate that the proposed framework demonstrates satisfactory performance in different scenarios:for the data replacement scenario,where real data are removed and replaced with generated data,the state estimator accuracy decreases only slightly;for the data enhancement scenario,the estimator accuracy is further improved.The estimation accuracy of SOH and SOC is as low as 0.69%and 0.58%root mean square error(RMSE)after applying the proposed framework.This framework provides a reliable method for enriching battery measurement data.It is a generalized framework capable of generating a variety of time series data.展开更多
基金supported by the National Natural Science Foundation of China (Grants Nos.61931004,62072250)the Talent Launch Fund of Nanjing University of Information Science and Technology (2020r061).
文摘Encrypted traffic classification has become a hot issue in network security research.The class imbalance problem of traffic samples often causes the deterioration of Machine Learning based classifier performance.Although the Generative Adversarial Network(GAN)method can generate new samples by learning the feature distribution of the original samples,it is confronted with the problems of unstable training andmode collapse.To this end,a novel data augmenting approach called Graph CWGAN-GP is proposed in this paper.The traffic data is first converted into grayscale images as the input for the proposed model.Then,the minority class data is augmented with our proposed model,which is built by introducing conditional constraints and a new distance metric in typical GAN.Finally,the classical deep learning model is adopted as a classifier to classify datasets augmented by the Condition GAN(CGAN),Wasserstein GAN-Gradient Penalty(WGAN-GP)and Graph CWGAN-GP,respectively.Compared with the state-of-the-art GAN methods,the Graph CWGAN-GP cannot only control the modes of the data to be generated,but also overcome the problem of unstable training and generate more realistic and diverse samples.The experimental results show that the classification precision,recall and F1-Score of theminority class in the balanced dataset augmented in this paper have improved by more than 2.37%,3.39% and 4.57%,respectively.
文摘Outcomes following peripheral nerve injury remain frustratingly poor. The reasons for this are multifactorial, although maintaining a growth permissive environment in the distal nerve stump following repair is arguably the most important. The optimal environment for axonal regeneration relies on the synthesis and release of many biochemical mediators that are temporally and spatially regulated with a high level of incompletely understood complexity. The Schwann cell(SC) has emerged as a key player in this process. Prolonged periods of distal nerve stump denervation, characteristic of large gaps and proximal injuries, have been associated with a reduction in SC number and ability to support regenerating axons. Cell based therapy offers a potential therapy for the improvement of outcomes following peripheral nerve reconstruction. Stem cells have the potential to increase the number of SCs and prolong their ability to support regeneration. They may also have the ability to rescue and replenish populations of chromatolytic and apoptotic neurons following axotomy. Finally, they can be used in non-physiologic ways to preserve injured tissues such as denervated muscle while neuronal ingrowth has not yet occurred. Aside from stem cell type, careful consideration must be given to differentiation status, how stem cells are supported following transplantation and how they will be delivered to the site of injury. It is the aim of this article to review current opinions on the strategies of stem cell based therapy for the augmentation of peripheral nerve regeneration.
基金supported by the Deanship of Graduate Studies and Scientific Research at Jouf University under grant No.DGSSR-2023-02-02116.
文摘When it comes to smart healthcare business systems,network-based intrusion detection systems are crucial for protecting the system and its networks from malicious network assaults.To protect IoMT devices and networks in healthcare and medical settings,our proposed model serves as a powerful tool for monitoring IoMT networks.This study presents a robust methodology for intrusion detection in Internet of Medical Things(IoMT)environments,integrating data augmentation,feature selection,and ensemble learning to effectively handle IoMT data complexity.Following rigorous preprocessing,including feature extraction,correlation removal,and Recursive Feature Elimi-nation(RFE),selected features are standardized and reshaped for deep learning models.Augmentation using the BAT algorithm enhances dataset variability.Three deep learning models,Transformer-based neural networks,self-attention Deep Convolutional Neural Networks(DCNNs),and Long Short-Term Memory(LSTM)networks,are trained to capture diverse data aspects.Their predictions form a meta-feature set for a subsequent meta-learner,which combines model strengths.Conventional classifiers validate meta-learner features for broad algorithm suitability.This comprehensive method demonstrates high accuracy and robustness in IoMT intrusion detection.Evaluations were conducted using two datasets:the publicly available WUSTL-EHMS-2020 dataset,which contains two distinct categories,and the CICIoMT2024 dataset,encompassing sixteen categories.Experimental results showcase the method’s exceptional performance,achieving optimal scores of 100%on the WUSTL-EHMS-2020 dataset and 99%on the CICIoMT2024.
基金This work was supported in part by the National Research Foundation of Korea(NRF)funded by the Ministry of Science and ICT(MSIT)Future Planning under Grant NRF-2020R1A2C2014336 and Grant NRF-2021R1A4A1029650.
文摘Android malware has evolved in various forms such as adware that continuously exposes advertisements,banking malware designed to access users’online banking accounts,and Short Message Service(SMS)malware that uses a Command&Control(C&C)server to send malicious SMS,intercept SMS,and steal data.By using many malicious strategies,the number of malware is steadily increasing.Increasing Android malware threats numerous users,and thus,it is necessary to detect malware quickly and accurately.Each malware has distinguishable characteristics based on its actions.Therefore,security researchers have tried to categorize malware based on their behaviors by conducting the familial analysis which can help analysists to reduce the time and cost for analyzing malware.However,those studies algorithms typically used imbalanced,well-labeled open-source dataset,and thus,it is very difficult to classify some malware families which only have a few number of malware.To overcome this challenge,previous data augmentation studies augmented data by visualizing malicious codes and used them for malware analysis.However,visualization of malware can result in misclassifications because the behavior information of the malware could be compromised.In this study,we propose an android malware familial analysis system based on a data augmentation method that preserves malware behaviors to create an effective multi-class classifier for malware family analysis.To this end,we analyze malware and use Application Programming Interface(APIs)and permissions that can reflect the behavior of malware as features.By using these features,we augment malware dataset to enable effective malware detection while preserving original malicious behaviors.Our evaluation results demonstrate that,when a model is created by using only the augmented data,a macro-F1 score of 0.65 and accuracy of 0.63%.On the other hand,when the augmented data and original malware are used together,the evaluation results show that a macro-F1 score of 0.91 and an accuracy of 0.99%.
基金supported by the National Natural Science Foundation of China(Grant Nos.U20B2049,U21B2018 and 62302344).
文摘Adversarial training has been widely considered the most effective defense against adversarial attacks.However,recent studies have demonstrated that a large discrepancy exists in the class-wise robustness of adversarial training,leading to two potential issues:firstly,the overall robustness of a model is compromised due to the weakest class;and secondly,ethical concerns arising from unequal protection and biases,where certain societal demographic groups receive less robustness in defense mechanisms.Despite these issues,solutions to address the discrepancy remain largely underexplored.In this paper,we advance beyond existing methods that focus on class-level solutions.Our investigation reveals that hard examples,identified by higher cross-entropy values,can provide more fine-grained information about the discrepancy.Furthermore,we find that enhancing the diversity of hard examples can effectively reduce the robustness gap between classes.Motivated by these observations,we propose Fair Adversarial Training(FairAT)to mitigate the discrepancy of class-wise robustness.Extensive experiments on various benchmark datasets and adversarial attacks demonstrate that FairAT outperforms state-of-the-art methods in terms of both overall robustness and fairness.For a WRN-28-10 model trained on CIFAR10,FairAT improves the average and worst-class robustness by 2.13%and 4.50%,respectively.
基金supported by National Natural Science Foundation of China:Space-based occultation detection with ground-based GNSS atmospheric horizontal gradient model(41904033).
文摘The satellite-based augmentation system(SBAS)provides differential and integrity augmentation services for life safety fields of aviation and navigation.However,the signal structure of SBAS is public,which incurs a risk of spoofing attacks.To improve the anti-spoofing capability of the SBAS,European Union and the United States conduct research on navigation message authentication,and promote the standardization of SBAS message authentication.For the development of Beidou satellite-based augmentation system(BDSBAS),this paper proposes navigation message authentication based on the Chinese commercial cryptographic standards.Firstly,this paper expounds the architecture and principles of the SBAS message authentication,and then carries out the design of timed efficient streaming losstolerant authentication scheme(TESLA)and elliptic curve digital signature algorithm(ECDSA)authentication schemes based on Chinese commercial cryptographic standards,message arrangement and the design of over-the-air rekeying(OTAR)message.Finally,this paper conducts a theoretical analysis of the time between authentications(TBA)and maximum authentication latency(MAL)for L5 TESLA-I and L5 ECDSA-Q,and further simulates the reception time of OTAR message,TBA and MAL from the aspects of OTAR message weight and demodulation error rate.The simulation results can provide theoretical supports for the standardization of BDSBAS message authentication.
基金supported by the project“Romanian Hub for Artificial Intelligence-HRIA”,Smart Growth,Digitization and Financial Instruments Program,2021–2027,MySMIS No.334906.
文摘Objective expertise evaluation of individuals,as a prerequisite stage for team formation,has been a long-term desideratum in large software development companies.With the rapid advancements in machine learning methods,based on reliable existing data stored in project management tools’datasets,automating this evaluation process becomes a natural step forward.In this context,our approach focuses on quantifying software developer expertise by using metadata from the task-tracking systems.For this,we mathematically formalize two categories of expertise:technology-specific expertise,which denotes the skills required for a particular technology,and general expertise,which encapsulates overall knowledge in the software industry.Afterward,we automatically classify the zones of expertise associated with each task a developer has worked on using Bidirectional Encoder Representations from Transformers(BERT)-like transformers to handle the unique characteristics of project tool datasets effectively.Finally,our method evaluates the proficiency of each software specialist across already completed projects from both technology-specific and general perspectives.The method was experimentally validated,yielding promising results.
基金partially supported by the National Natural Science Foundation of China(62075137)the Guangdong Basic and Applied Basic Research Foundation(2023A1515140161)+3 种基金the Guangxi Natural Science Foundation of China(2021JJB 110003)the Dongguan Science and Technology of Social Development Program(20231800936312)the high-level talent program of Dongguan University of Technology(No.221110080)the Sanming Project of Medicine in Shenzhen(No.SZSM202103014).
文摘Many spore-forming Bacillus species can cause serious human diseases,because of accidental Bacillusspore infection.Thus,developing an identification strategy with both high sensitivity and specificity is greatly in demand.In this work,we proposed a novel approach named multi-head self-attention mechanism-guided neural network Raman platform to identify living Bacillus spores within a single-cell resolution.The multi-head self-attention mechanism-guided neural network Raman platform was created by combining single-cell Raman spectroscopy,convolutional neural network(CNN),and multi-head self-attention mechanism.To address the limited size of the original spectra dataset,Gaussian noise-based spectra augmentation was employed to increase the number of single-cell Raman spectra datasets for CNN training.Owing to the assistance of both spectra augmentation and multi-head self-attention mechanism,the obtained prediction accuracy of five Bacillus spore species was further improved from 92.29±0.82%to 99.43±0.15%.To figure out the spectra differences covered by the multi-head self-attention mechanism-guided CNN,the relative classification weight from typical Raman bands was visualized via multi-head self-attention mechanism curve.In the process of spectra augmentation from 0 to 1000,the distribution of relative classification weight varied from a discrete state to a more concentrated phase.More importantly,these highlighted four Raman bands(1017,1449,1576,and 1660 cm^(-1))were assigned large weights,showing that the spectra differences in the Raman bands produced the largest contribution to prediction accuracy.It can be foreseen that,our proposed sorting platform has great potential in accurately identifying Bacillus and its related genera species at a single-cell level.
基金supported by the National Natural Science Foundation of China under Grant Nos.61836007 and 61772354.
文摘Due to the small size of the annotated corpora and the sparsity of the event trigger words, the event coreference resolver cannot capture enough event semantics, especially the trigger semantics, to identify coreferential event mentions. To address the above issues, this paper proposes a trigger semantics augmentation mechanism to boost event coreference resolution. First, this mechanism performs a trigger-oriented masking strategy to pre-train a BERT (Bidirectional Encoder Representations from Transformers)-based encoder (Trigger-BERT), which is fine-tuned on a large-scale unlabeled dataset Gigaword. Second, it combines the event semantic relations from the Trigger-BERT encoder with the event interactions from the soft-attention mechanism to resolve event coreference. Experimental results on both the KBP2016 and KBP2017 datasets show that our proposed model outperforms several state-of-the-art baselines.
文摘Lung cancer continues to be a leading cause of cancer-related deaths worldwide,emphasizing the critical need for improved diagnostic techniques.Early detection of lung tumors significantly increases the chances of successful treatment and survival.However,current diagnostic methods often fail to detect tumors at an early stage or to accurately pinpoint their location within the lung tissue.Single-model deep learning technologies for lung cancer detection,while beneficial,cannot capture the full range of features present in medical imaging data,leading to incomplete or inaccurate detection.Furthermore,it may not be robust enough to handle the wide variability in medical images due to different imaging conditions,patient anatomy,and tumor characteristics.To overcome these disadvantages,dual-model or multi-model approaches can be employed.This research focuses on enhancing the detection of lung cancer by utilizing a combination of two learning models:a Convolutional Neural Network(CNN)for categorization and the You Only Look Once(YOLOv8)architecture for real-time identification and pinpointing of tumors.CNNs automatically learn to extract hierarchical features from raw image data,capturing patterns such as edges,textures,and complex structures that are crucial for identifying lung cancer.YOLOv8 incorporates multiscale feature extraction,enabling the detection of tumors of varying sizes and scales within a single image.This is particularly beneficial for identifying small or irregularly shaped tumors that may be challenging to detect.Furthermore,through the utilization of cutting-edge data augmentation methods,such as Deep Convolutional Generative Adversarial Networks(DCGAN),the suggested approach can handle the issue of limited data and boost the models’ability to learn from diverse and comprehensive datasets.The combined method not only improved accuracy and localization but also ensured efficient real-time processing,which is crucial for practical clinical applications.The CNN achieved an accuracy of 97.67%in classifying lung tissues into healthy and cancerous categories.The YOLOv8 model achieved an Intersection over Union(IoU)score of 0.85 for tumor localization,reflecting high precision in detecting and marking tumor boundaries within the images.Finally,the incorporation of synthetic images generated by DCGAN led to a 10%improvement in both the CNN classification accuracy and YOLOv8 detection performance.
基金National Natural Science Foundation of China(62171305,62405206,62004135,62001317,62111530301)Natural Science Foundation of Jiangsu Province(BK20240778,BK20241917)+3 种基金State Key Laboratory of Advanced Optical Communication Systems and Networks,China(2023GZKF08)China Postdoctoral Science Foundation(2024M752314)Postdoctoral Fellowship Program of CPSF(GZC20231883)Innovative and Entrepreneurial Talent Program of Jiangsu Province(JSSCRC2021527).
文摘Photonic platforms are gradually emerging as a promising option to encounter the ever-growing demand for artificial intelligence,among which photonic time-delay reservoir computing(TDRC)is widely anticipated.While such a computing paradigm can only employ a single photonic device as the nonlinear node for data processing,the performance highly relies on the fading memory provided by the delay feedback loop(FL),which sets a restriction on the extensibility of physical implementation,especially for highly integrated chips.Here,we present a simplified photonic scheme for more flexible parameter configurations leveraging the designed quasi-convolution coding(QC),which completely gets rid of the dependence on FL.Unlike delay-based TDRC,encoded data in QC-based RC(QRC)enables temporal feature extraction,facilitating augmented memory capabilities.Thus,our proposed QRC is enabled to deal with time-related tasks or sequential data without the implementation of FL.Furthermore,we can implement this hardware with a low-power,easily integrable vertical-cavity surface-emitting laser for high-performance parallel processing.We illustrate the concept validation through simulation and experimental comparison of QRC and TDRC,wherein the simpler-structured QRC outperforms across various benchmark tasks.Our results may underscore an auspicious solution for the hardware implementation of deep neural networks.
文摘Objective The study of medicine formulas is a core component of traditional Chinese medicine(TCM),yet traditional learning methods often lack interactivity and contextual understanding,making it challenging for beginners to grasp the intricate composition rules of formulas.To address this gap,we introduce Formula-S,a situated visualization method for TCM formula learning in augmented reality(AR)and evaluate its performance.This study aims to evaluate the effectiveness of Formula-S in enhancing TCM formula learning for beginners by comparing it with traditional text-based formula learning and web-based visualization.Methods Formula-S is an interactive AR tool designed for TCM formula learning,featuring three modes(3D,Web,and Table).The dataset included TCM formulas and herb properties extracted from authoritative references,including textbook and the SymMap database.In Formula-S,the hierarchical visualization of the formulas as herbal medicine compositions,is linked to the multidimensional herb attribute visualization and embedded in the real world,where real herb samples are presented.To evaluate its effectiveness,a controlled study(n=30)was conducted.Participants who had no formal TCM knowledge were tasked with herbal medicine identification,formula composition,and recognition.In the study,participants interacted with the AR tool through HoloLens 2.Data were collected on both task performance(accuracy and response time)and user experience,with a focus on task efficiency,accuracy,and user preference across the different learning modes.Results The situated visualization method of Formula-S had comparable accuracy to other methods but shorter response time for herbal formula learning tasks.Regarding user experience,our new approach demonstrated the highest system usability and lowest task load,effectively reducing cognitive load and allowing users to complete tasks with greater ease and efficiency.Participants reported that Formula-S enhanced their learning experience through its intuitive interface and immersive AR environment,suggesting this approach offers usability advantages for TCM education.Conclusions The situated visualization method in Formula-S offers more efficient and accurate searching capabilities compared to traditional and web-based methods.Additionally,it provides superior contextual understanding of TCM formulas,making it a promising new solution for TCM learning.
文摘The integration of image analysis through deep learning(DL)into rock classification represents a significant leap forward in geological research.While traditional methods remain invaluable for their expertise and historical context,DL offers a powerful complement by enhancing the speed,objectivity,and precision of the classification process.This research explores the significance of image data augmentation techniques in optimizing the performance of convolutional neural networks(CNNs)for geological image analysis,particularly in the classification of igneous,metamorphic,and sedimentary rock types from rock thin section(RTS)images.This study primarily focuses on classic image augmentation techniques and evaluates their impact on model accuracy and precision.Results demonstrate that augmentation techniques like Equalize significantly enhance the model's classification capabilities,achieving an F1-Score of 0.9869 for igneous rocks,0.9884 for metamorphic rocks,and 0.9929 for sedimentary rocks,representing improvements compared to the baseline original results.Moreover,the weighted average F1-Score across all classes and techniques is 0.9886,indicating an enhancement.Conversely,methods like Distort lead to decreased accuracy and F1-Score,with an F1-Score of 0.949 for igneous rocks,0.954 for metamorphic rocks,and 0.9416 for sedimentary rocks,exacerbating the performance compared to the baseline.The study underscores the practicality of image data augmentation in geological image classification and advocates for the adoption of DL methods in this domain for automation and improved results.The findings of this study can benefit various fields,including remote sensing,mineral exploration,and environmental monitoring,by enhancing the accuracy of geological image analysis both for scientific research and industrial applications.
文摘Breast cancer is one of the major causes of deaths in women.However,the early diagnosis is important for screening and control the mortality rate.Thus for the diagnosis of breast cancer at the early stage,a computer-aided diagnosis system is highly required.Ultrasound is an important examination technique for breast cancer diagnosis due to its low cost.Recently,many learning-based techniques have been introduced to classify breast cancer using breast ultrasound imaging dataset(BUSI)datasets;however,the manual handling is not an easy process and time consuming.The authors propose an EfficientNet-integrated ResNet deep network and XAI-based framework for accurately classifying breast cancer(malignant and benign).In the initial step,data augmentation is performed to increase the number of training samples.For this purpose,three-pixel flip mathematical equations are introduced:horizontal,vertical,and 90°.Later,two pretrained deep learning models were employed,skipped some layers,and fine-tuned.Both fine-tuned models are later trained using a deep transfer learning process and extracted features from the deeper layer.Explainable artificial intelligence-based analysed the performance of trained models.After that,a new feature selection technique is proposed based on the cuckoo search algorithm called cuckoo search controlled standard error mean.This technique selects the best features and fuses using a new parallel zeropadding maximum correlated coefficient features.In the end,the selection algorithm is applied again to the fused feature vector and classified using machine learning algorithms.The experimental process of the proposed framework is conducted on a publicly available BUSI and obtained 98.4%and 98%accuracy in two different experiments.Comparing the proposed framework is also conducted with recent techniques and shows improved accuracy.In addition,the proposed framework was executed less than the original deep learning models.
基金Postgraduate Innovation Top notch Talent Training Project of Hunan Province,Grant/Award Number:CX20220045Scientific Research Project of National University of Defense Technology,Grant/Award Number:22-ZZCX-07+2 种基金New Era Education Quality Project of Anhui Province,Grant/Award Number:2023cxcysj194National Natural Science Foundation of China,Grant/Award Numbers:62201597,62205372,1210456foundation of Hefei Comprehensive National Science Center,Grant/Award Number:KY23C502。
文摘Large-scale point cloud datasets form the basis for training various deep learning networks and achieving high-quality network processing tasks.Due to the diversity and robustness constraints of the data,data augmentation(DA)methods are utilised to expand dataset diversity and scale.However,due to the complex and distinct characteristics of LiDAR point cloud data from different platforms(such as missile-borne and vehicular LiDAR data),directly applying traditional 2D visual domain DA methods to 3D data can lead to networks trained using this approach not robustly achieving the corresponding tasks.To address this issue,the present study explores DA for missile-borne LiDAR point cloud using a Monte Carlo(MC)simulation method that closely resembles practical application.Firstly,the model of multi-sensor imaging system is established,taking into account the joint errors arising from the platform itself and the relative motion during the imaging process.A distortion simulation method based on MC simulation for augmenting missile-borne LiDAR point cloud data is proposed,underpinned by an analysis of combined errors between different modal sensors,achieving high-quality augmentation of point cloud data.The effectiveness of the proposed method in addressing imaging system errors and distortion simulation is validated using the imaging scene dataset constructed in this paper.Comparative experiments between the proposed point cloud DA algorithm and the current state-of-the-art algorithms in point cloud detection and single object tracking tasks demonstrate that the proposed method can improve the network performance obtained from unaugmented datasets by over 17.3%and 17.9%,surpassing SOTA performance of current point cloud DA algorithms.
文摘Solar cell defects exhibit significant variations and multiple types,with some defect data being difficult to acquire or having small scales,posing challenges in terms of small sample and small target in defect detection for solar cells.In order to address this issue,this paper proposes a multi-step approach for detecting the complex defects of solar cells.First,individual cell plates are extracted from electroluminescence images for block-by-block detection.Then,StyleGAN2-Ada is utilized for generative adversarial networks data augmentation to expand the number of defect samples in small sample defects.Finally,the fake dataset is combined with real dataset,and the improved YOLOv5 model is trained on this mixed dataset.Experimental results demonstrate that the proposed method achieves a superior performance in detecting the defects with small sample and small target,with the final recall rate reaching 99.7%,an increase of 3.9% compared with the unimproved model.Additionally,the precision and mean average precision are increased by 3.4% and 3.5%,respectively.Moreover,the experiments demonstrate that the improved network training on the mixed dataset can effectively enhance the detection performance of the model.The combination of these approaches significantly improves the network’s ability to detect solar cell defects.
文摘Phantom limb pain(PLP)is not only a physical pain experience but also poses a significant challenge to mental health and quality of life.Currently,the mechanism of PLP treatment is still unclear,and there are many methods with varying effects.This article starts with the application research of extended reality technology in PLP treatment,through describing the application of its branch technologies(virtual reality,augmented reality,and mixed reality technology),to lay the foundation for subsequent research,in the hope of finding advanced and effective treatment methods,and providing a basis for future product transformation.
基金Natural Science Foundation of Zhejiang Province,Grant/Award Number:LY23F020025Science and Technology Commissioner Program of Huzhou,Grant/Award Number:2023GZ42Sichuan Provincial Science and Technology Support Program,Grant/Award Numbers:2023ZHCG0005,2023ZHCG0008。
文摘Data augmentation plays an important role in training deep neural model by expanding the size and diversity of the dataset.Initially,data augmentation mainly involved some simple transformations of images.Later,in order to increase the diversity and complexity of data,more advanced methods appeared and evolved to sophisticated generative models.However,these methods required a mass of computation of training or searching.In this paper,a novel training-free method that utilises the Pre-Trained Segment Anything Model(SAM)model as a data augmentation tool(PTSAM-DA)is proposed to generate the augmented annotations for images.Without the need for training,it obtains prompt boxes from the original annotations and then feeds the boxes to the pre-trained SAM to generate diverse and improved annotations.In this way,annotations are augmented more ingenious than simple manipulations without incurring huge computation for training a data augmentation model.Multiple comparative experiments on three datasets are conducted,including an in-house dataset,ADE20K and COCO2017.On this in-house dataset,namely Agricultural Plot Segmentation Dataset,maximum improvements of 3.77%and 8.92%are gained in two mainstream metrics,mIoU and mAcc,respectively.Consequently,large vision models like SAM are proven to be promising not only in image segmentation but also in data augmentation.
基金funded by the Bavarian State Ministry of Science,Research and Art(Grant number:H.2-F1116.WE/52/2)。
文摘In order to address the widespread data shortage problem in battery research,this paper proposes a generative adversarial network model that combines it with deep convolutional networks,the Wasserstein distance,and the gradient penalty to achieve data augmentation.To lower the threshold for implementing the proposed method,transfer learning is further introduced.The W-DC-GAN-GP-TL framework is thereby formed.This framework is evaluated on 3 different publicly available datasets to judge the quality of generated data.Through visual comparisons and the examination of two visualization methods(probability density function(PDF)and principal component analysis(PCA)),it is demonstrated that the generated data is hard to distinguish from the real data.The application of generated data for training a battery state model using transfer learning is further evaluated.Specifically,Bi-GRU-based and Transformer-based methods are implemented on 2 separate datasets for estimating state of health(SOH)and state of charge(SOC),respectively.The results indicate that the proposed framework demonstrates satisfactory performance in different scenarios:for the data replacement scenario,where real data are removed and replaced with generated data,the state estimator accuracy decreases only slightly;for the data enhancement scenario,the estimator accuracy is further improved.The estimation accuracy of SOH and SOC is as low as 0.69%and 0.58%root mean square error(RMSE)after applying the proposed framework.This framework provides a reliable method for enriching battery measurement data.It is a generalized framework capable of generating a variety of time series data.
