To address the issue of scarce labeled samples and operational condition variations that degrade the accuracy of fault diagnosis models in variable-condition gearbox fault diagnosis,this paper proposes a semi-supervis...To address the issue of scarce labeled samples and operational condition variations that degrade the accuracy of fault diagnosis models in variable-condition gearbox fault diagnosis,this paper proposes a semi-supervised masked contrastive learning and domain adaptation(SSMCL-DA)method for gearbox fault diagnosis under variable conditions.Initially,during the unsupervised pre-training phase,a dual signal augmentation strategy is devised,which simultaneously applies random masking in the time domain and random scaling in the frequency domain to unlabeled samples,thereby constructing more challenging positive sample pairs to guide the encoder in learning intrinsic features robust to condition variations.Subsequently,a ConvNeXt-Transformer hybrid architecture is employed,integrating the superior local detail modeling capacity of ConvNeXt with the robust global perception capability of Transformer to enhance feature extraction in complex scenarios.Thereafter,a contrastive learning model is constructed with the optimization objective of maximizing feature similarity across different masked instances of the same sample,enabling the extraction of consistent features from multiple masked perspectives and reducing reliance on labeled data.In the final supervised fine-tuning phase,a multi-scale attention mechanism is incorporated for feature rectification,and a domain adaptation module combining Local Maximum Mean Discrepancy(LMMD)with adversarial learning is proposed.This module embodies a dual mechanism:LMMD facilitates fine-grained class-conditional alignment,compelling features of identical fault classes to converge across varying conditions,while the domain discriminator utilizes adversarial training to guide the feature extractor toward learning domain-invariant features.Working in concert,they markedly diminish feature distribution discrepancies induced by changes in load,rotational speed,and other factors,thereby boosting the model’s adaptability to cross-condition scenarios.Experimental evaluations on the WT planetary gearbox dataset and the Case Western Reserve University(CWRU)bearing dataset demonstrate that the SSMCL-DA model effectively identifies multiple fault classes in gearboxes,with diagnostic performance substantially surpassing that of conventional methods.Under cross-condition scenarios,the model attains fault diagnosis accuracies of 99.21%for the WT planetary gearbox and 99.86%for the bearings,respectively.Furthermore,the model exhibits stable generalization capability in cross-device settings.展开更多
In the field of intelligent air combat,real-time and accurate recognition of within-visual-range(WVR)maneuver actions serves as the foundational cornerstone for constructing autonomous decision-making systems.However,...In the field of intelligent air combat,real-time and accurate recognition of within-visual-range(WVR)maneuver actions serves as the foundational cornerstone for constructing autonomous decision-making systems.However,existing methods face two major challenges:traditional feature engineering suffers from insufficient effective dimensionality in the feature space due to kinematic coupling,making it difficult to distinguish essential differences between maneuvers,while end-to-end deep learning models lack controllability in implicit feature learning and fail to model high-order long-range temporal dependencies.This paper proposes a trajectory feature pre-extraction method based on a Long-range Masked Autoencoder(LMAE),incorporating three key innovations:(1)Random Fragment High-ratio Masking(RFH-Mask),which enforces the model to learn long-range temporal correlations by masking 80%of trajectory data while retaining continuous fragments;(2)Kalman Filter-Guided Objective Function(KFG-OF),integrating trajectory continuity constraints to align the feature space with kinematic principles;and(3)Two-stage Decoupled Architecture,enabling efficient and controllable feature learning through unsupervised pre-training and frozen-feature transfer.Experimental results demonstrate that LMAE significantly improves the average recognition accuracy for 20-class maneuvers compared to traditional end-to-end models,while significantly accelerating convergence speed.The contributions of this work lie in:introducing high-masking-rate autoencoders into low-informationdensity trajectory analysis,proposing a feature engineering framework with enhanced controllability and efficiency,and providing a novel technical pathway for intelligent air combat decision-making systems.展开更多
Recently,several PC oracle based side-channel attacks have been proposed against Kyber.However,most of them focus on unprotected implementations and masking is considered as a counter-measure.In this study,we extend P...Recently,several PC oracle based side-channel attacks have been proposed against Kyber.However,most of them focus on unprotected implementations and masking is considered as a counter-measure.In this study,we extend PC oracle based side-channel attacks to the second-order scenario and successfully conduct key-recovery attacks on the first-order masked Kyber.Firstly,we analyze the potential joint information leakage.Inspired by the binary PC oracle based attack proposed by Qin et al.at Asiacrypt 2021,we identify the 1-bit leakage scenario in the masked Keccak implementation.Moreover,we modify the ciphertexts construction described by Tanaka et al.at CHES 2023,extending the leakage scenario from 1-bit to 32-bit.With the assistance of TVLA,we validate these leakages through experiments.Secondly,for these two scenarios,we construct a binary PC oracle based on t-test and a multiple-valued PC oracle based on neural networks.Furthermore,we conduct practical side-channel attacks on masked Kyber by utilizing our oracles,with the implementation running on an ARM Cortex-M4 microcontroller.The demonstrated attacks require a minimum of 15788 and 648 traces to fully recover the key of Kyber768 in the 1-bit leakage scenario and the 32-bit leakage scenario,respectively.Our analysis may also be extended to attack other post-quantum schemes that use the same masked hash function.Finally,we apply the shuffling strategy to the first-order masked imple-mentation of the Kyber and perform leakage tests.Experimental results show that the combination strategy of shuffling and masking can effectively resist our proposed attacks.展开更多
Purpose:Mo ve recognition in scientific abstracts is an NLP task of classifying sentences of the abstracts into different types of language units.To improve the performance of move recognition in scientific abstracts,...Purpose:Mo ve recognition in scientific abstracts is an NLP task of classifying sentences of the abstracts into different types of language units.To improve the performance of move recognition in scientific abstracts,a novel model of move recognition is proposed that outperforms the BERT-based method.Design/methodology/approach:Prevalent models based on BERT for sentence classification often classify sentences without considering the context of the sentences.In this paper,inspired by the BERT masked language model(MLM),we propose a novel model called the masked sentence model that integrates the content and contextual information of the sentences in move recognition.Experiments are conducted on the benchmark dataset PubMed 20K RCT in three steps.Then,we compare our model with HSLN-RNN,BERT-based and SciBERT using the same dataset.Findings:Compared with the BERT-based and SciBERT models,the F1 score of our model outperforms them by 4.96%and 4.34%,respectively,which shows the feasibility and effectiveness of the novel model and the result of our model comes closest to the state-of-theart results of HSLN-RNN at present.Research limitations:The sequential features of move labels are not considered,which might be one of the reasons why HSLN-RNN has better performance.Our model is restricted to dealing with biomedical English literature because we use a dataset from PubMed,which is a typical biomedical database,to fine-tune our model.Practical implications:The proposed model is better and simpler in identifying move structures in scientific abstracts and is worthy of text classification experiments for capturing contextual features of sentences.Originality/value:T he study proposes a masked sentence model based on BERT that considers the contextual features of the sentences in abstracts in a new way.The performance of this classification model is significantly improved by rebuilding the input layer without changing the structure of neural networks.展开更多
Deep convolutional neural networks(DCNNs)are widely used in content-based image retrieval(CBIR)because of the advantages in image feature extraction.However,the training of deep neural networks requires a large number...Deep convolutional neural networks(DCNNs)are widely used in content-based image retrieval(CBIR)because of the advantages in image feature extraction.However,the training of deep neural networks requires a large number of labeled data,which limits the application.Self-supervised learning is a more general approach in unlabeled scenarios.A method of fine-tuning feature extraction networks based on masked learning is proposed.Masked autoencoders(MAE)are used in the fine-tune vision transformer(ViT)model.In addition,the scheme of extracting image descriptors is discussed.The encoder of the MAE uses the ViT to extract global features and performs self-supervised fine-tuning by reconstructing masked area pixels.The method works well on category-level image retrieval datasets with marked improvements in instance-level datasets.For the instance-level datasets Oxford5k and Paris6k,the retrieval accuracy of the base model is improved by 7%and 17%compared to that of the original model,respectively.展开更多
We consider a series system of two independent and non-identical components which have different BurrⅫ distributed lifetime.The maximum likelihood and Bayes estimators of the parameters of the system＇s components ar...We consider a series system of two independent and non-identical components which have different BurrⅫ distributed lifetime.The maximum likelihood and Bayes estimators of the parameters of the system＇s components are obtained based on masked system life test data.The conclusion is that the Bayes estimates are better than the maximum likelihood estimates in the sense of having smaller mean squared errors.展开更多
Significant advancements have beenwitnessed in visual tracking applications leveragingViT in recent years,mainly due to the formidablemodeling capabilities of Vision Transformer(ViT).However,the strong performance of ...Significant advancements have beenwitnessed in visual tracking applications leveragingViT in recent years,mainly due to the formidablemodeling capabilities of Vision Transformer(ViT).However,the strong performance of such trackers heavily relies on ViT models pretrained for long periods,limitingmore flexible model designs for tracking tasks.To address this issue,we propose an efficient unsupervised ViT pretraining method for the tracking task based on masked autoencoders,called TrackMAE.During pretraining,we employ two shared-parameter ViTs,serving as the appearance encoder and motion encoder,respectively.The appearance encoder encodes randomly masked image data,while the motion encoder encodes randomly masked pairs of video frames.Subsequently,an appearance decoder and a motion decoder separately reconstruct the original image data and video frame data at the pixel level.In this way,ViT learns to understand both the appearance of images and the motion between video frames simultaneously.Experimental results demonstrate that ViT-Base and ViT-Large models,pretrained with TrackMAE and combined with a simple tracking head,achieve state-of-the-art(SOTA)performance without additional design.Moreover,compared to the currently popular MAE pretraining methods,TrackMAE consumes only 1/5 of the training time,which will facilitate the customization of diverse models for tracking.For instance,we additionally customize a lightweight ViT-XS,which achieves SOTA efficient tracking performance.展开更多
Since the outbreak of Coronavirus Disease 2019(COVID-19),people are recommended to wear facial masks to limit the spread of the virus.Under the circumstances,traditional face recognition technologies cannot achieve sa...Since the outbreak of Coronavirus Disease 2019(COVID-19),people are recommended to wear facial masks to limit the spread of the virus.Under the circumstances,traditional face recognition technologies cannot achieve satisfactory results.In this paper,we propose a face recognition algorithm that combines the traditional features and deep features of masked faces.For traditional features,we extract Local Binary Pattern(LBP),Scale-Invariant Feature Transform(SIFT)and Histogram of Oriented Gradient(HOG)features from the periocular region,and use the Support Vector Machines(SVM)classifier to perform personal identification.We also propose an improved Convolutional Neural Network(CNN)model Angular Visual Geometry Group Network(A-VGG)to learn deep features.Then we use the decision-level fusion to combine the four features.Comprehensive experiments were carried out on databases of real masked faces and simulated masked faces,including frontal and side faces taken at different angles.Images with motion blur were also tested to evaluate the robustness of the algorithm.Besides,the experiment of matching a masked face with the corresponding full face is accomplished.The experimental results show that the proposed algorithm has state-of-the-art performance in masked face recognition,and the periocular region has rich biological features and high discrimination.展开更多
Under Type-Ⅱ progressively hybrid censoring, this paper discusses statistical inference and optimal design on stepstress partially accelerated life test for hybrid system in presence of masked data. It is assumed tha...Under Type-Ⅱ progressively hybrid censoring, this paper discusses statistical inference and optimal design on stepstress partially accelerated life test for hybrid system in presence of masked data. It is assumed that the lifetime of the component in hybrid systems follows independent and identical modified Weibull distributions. The maximum likelihood estimations(MLEs)of the unknown parameters, acceleration factor and reliability indexes are derived by using the Newton-Raphson algorithm. The asymptotic variance-covariance matrix and the approximate confidence intervals are obtained based on normal approximation to the asymptotic distribution of MLEs of model parameters. Moreover,two bootstrap confidence intervals are constructed by using the parametric bootstrap method. The optimal time of changing stress levels is determined under D-optimality and A-optimality criteria.Finally, the Monte Carlo simulation study is carried out to illustrate the proposed procedures.展开更多
The fat production rate in adult healthy masked civet(Paguma lavata) and nutria (Myocaster coypus) oil were measured. The values of iodine. saponification and acid PH, composition of fstty acids of grease were analyze...The fat production rate in adult healthy masked civet(Paguma lavata) and nutria (Myocaster coypus) oil were measured. The values of iodine. saponification and acid PH, composition of fstty acids of grease were analyzed both chemically and by apparatus. The results showed that acid PH, iodine value, saponification value,and unsaturation point are 1.887 and 0.784, 53.90 and 48.32, 98.80 and 100.23. and 60.05% and 58.85% are respectively for masked civet's fat and nutria's oil. Both of masked civet's fat and nutria's oil contain a little of Eicosatetraenoic acid (C-20;4), which is of great significance in nutrition and metabolism for human body. The analysis results indicate that masked civet's oil is similar to nutria's oil in iodine value, saponification value and unsaturation point. Both masked civet's fat and nutria's oil are steady and have highly nutrition. They can be widely exploited and utilized in health protection and cosmetics made industry.展开更多
Background Hypertension is the main risk factor for cardiovascular diseases, affecting more than half the elderly population. It is essential to know if they have proper control of hypertension. The aim of this study ...Background Hypertension is the main risk factor for cardiovascular diseases, affecting more than half the elderly population. It is essential to know if they have proper control of hypertension. The aim of this study was to identify the associated factors to masked uncon- trolled hypertension and false uncontrolled hypertension in older patients. Methods Two-hundred seventy-three individuals （70.1±6.7 years-old） had blood pressure （BP） measured at the office and by ambulatory BP monitoring （ABPM）, with the definition of controlled group （C）, individuals with high office BP and adequate ABPM, called white-coat effect group （WCE）, uncontrolled （UC）, and subjects with ap- propriate office BP and elevated ABPM denominated masked effect group （ME）. Age, body mass index, diabetes, pulse pressure （PP） and BP dipping during sleep were evaluated （Kruskal-Wallis test and logistic regression models）. Results Age was higher in UC than in C and ME （P 〈 0.01）, and 24-h ABPM PP was lower in C （48± 7 mmHg） and WCE （51±6 mmHg） than in UC （67±12 mmHg） and ME （59±8 mmHg） （P 〈 0.01）. Sleep systolic BP dipping was lower in ME than in C （P = 0.03）. Female gender was associated with a greater chance of being of ME group, which showed a higher PP and lower BP dipping during sleep. Conclusions In older individuals, office BP measure- ments did not allow the detection of associated factors that would permit to differentiate WCE from UC group and C from ME group. ABPM favored the identification of a higher PP and a lower BP dipping during sleep in the masked effect and uncontrolled groups.展开更多
Objective:To compare the risk of target organ damage in masked hypertension(MH)and sustained hypertension(SH).Methods:A systematic review and meta-analysis was performed.A search of PubMed,Embase,and the Cochrane Libr...Objective:To compare the risk of target organ damage in masked hypertension(MH)and sustained hypertension(SH).Methods:A systematic review and meta-analysis was performed.A search of PubMed,Embase,and the Cochrane Library of relevant case-control studies was performed from inception to December 2019,and articles on MH and SH selected according to the inclusion criteria were analyzed.The primary end point was target organ damage in the heart.The secondary end points were target organ damage in the kidneys and blood vessels.Results:Seventeen studies that met the screening criteria were included in the meta-analysis.Compared with the SH group,in the MH group carotid intima-media thickness(IMT)and E/A ratio were signifi cantly greater and the prevalence of left ventricular remodeling and the pulse wave velocity were signifi cantly lower.Other indicators in the heart,kidneys,and blood vessels were not statistically different between the two groups.IMT:P=0.01,E/A ratio:P=0.01,prevalence of left ventricular remodeling:P=0.02,pulse wave velocity:P=0.01.Conclusion:Our study has shown that MH may have almost the same degree of target organ damage as SH,so clinicians may need to consider target organ damage.展开更多
The purpose of this study was to cover the bitter taste of arbidol hydrochloride(ARB)and develop dry suspension with combination of solid dispersion and flavors.Taste masking was successfully done by solid dispersion ...The purpose of this study was to cover the bitter taste of arbidol hydrochloride(ARB)and develop dry suspension with combination of solid dispersion and flavors.Taste masking was successfully done by solid dispersion using octadecanol as the carrier by fusion method.Suspending agents,carriers and other excipients were selected.Differential scanning calorimetry(DSC)and Fourier transform infrared spectroscopy(FTIR)were performed to identify the physicochemical interaction between drug and carrier,DSC analysis indicated that ARB was amorphous in the solid dispersion,FTIR spectroscopy showed no interaction between drug and carrier.Taste masking was evaluated on six volunteers with a score of 4.9.The results demonstrated successful taste masking.Water was used to study the in vitro dissolution performance of the three formulations of commercial tablet,capsule and self-made suspension.The self-made suspension showed a lower and slower release,the insoluble carrier octadecanol blocked the drug dissolving from the solid dispersion.It was indicated from the primary stability study,the self-made suspensions were sensitive to high temperature,high humidity and strong light conditions,they should be stored in sealed containers away from heat,light and humidity.展开更多
Cefuroxime axetil(CA)is an ester prodrug of cefuroxime with an unpleasant taste when administrated orally.This work was to mask the bitter taste of CA and enhance its oral bioavailability.Dry suspensions were prepared...Cefuroxime axetil(CA)is an ester prodrug of cefuroxime with an unpleasant taste when administrated orally.This work was to mask the bitter taste of CA and enhance its oral bioavailability.Dry suspensions were prepared by means of wet granulation method and solid dispersion method.Binders,suspending agents and other compositions involved in the formulation were optimized.The differential scanning calorimetry(DSC)analysis indicated that CA was amorphous in the solid dispersion with stearic acid as the carrier,which contributed to an improvement of the dissolution rate.Taste evaluation was performed by three volunteers and taste masking was successfully achieved by the methods mentioned above.A pH 7.0 phosphate buffer was adopted to study the in vitro dissolution performance of the three formulations,i.e.,two self-made dry suspensions and the commercial one.With a better release characteristic and a satisfying taste masking ability,the solid dispersion suspension was selected as the optimal formulation for the further pharmacokinetic study in beagle dogs.The values of Cmax and AUC0e12 for the solid dispersion suspension were about 1.78-fold and 2.17-fold higher than these of reference suspension,respectively.The obtained results demonstrated that the solid dispersion can efficiently mask the bitter taste of CA and significantly enhance its oral bioavailability.展开更多
Civets are small mammals belonging to the family Viverridae.The masked palm civets(Paguma larvata)served as an intermediate host in the bat-to-human transmission of severe acute respiratory syndrome coronavirus(SARS-C...Civets are small mammals belonging to the family Viverridae.The masked palm civets(Paguma larvata)served as an intermediate host in the bat-to-human transmission of severe acute respiratory syndrome coronavirus(SARS-Co V)in 2003(Guan et al.,2003).Because of their unique role in the SARS outbreak,civets were suspected as a potential intermediate host of SARS-Co V-2.展开更多
Face mask detection has several applications,including real-time surveillance,biometrics,etc.Identifying face masks is also helpful for crowd control and ensuring people wear them publicly.With monitoring personnel,it...Face mask detection has several applications,including real-time surveillance,biometrics,etc.Identifying face masks is also helpful for crowd control and ensuring people wear them publicly.With monitoring personnel,it is impossible to ensure that people wear face masks;automated systems are a much superior option for face mask detection and monitoring.This paper introduces a simple and efficient approach for masked face detection.The architecture of the proposed approach is very straightforward;it combines deep learning and local binary patterns to extract features and classify themasmasked or unmasked.The proposed systemrequires hardware withminimal power consumption compared to state-of-the-art deep learning algorithms.Our proposed system maintains two steps.At first,this work extracted the local features of an image by using a local binary pattern descriptor,and then we used deep learning to extract global features.The proposed approach has achieved excellent accuracy and high performance.The performance of the proposed method was tested on three benchmark datasets:the realworld masked faces dataset(RMFD),the simulated masked faces dataset(SMFD),and labeled faces in the wild(LFW).Performancemetrics for the proposed technique weremeasured in terms of accuracy,precision,recall,and F1-score.Results indicated the efficiency of the proposed technique,providing accuracies of 99.86%,99.98%,and 100%for RMFD,SMFD,and LFW,respectively.Moreover,the proposed method outperformed state-of-the-art deep learning methods in the recent bibliography for the same problem under study and on the same evaluation datasets.展开更多
Masked autoencoders(MAEs)have emerged as a powerful self-supervised approach for point cloud analysis.Nevertheless,existing methods often separately focus on global structures or multi-scale features,ignoring their co...Masked autoencoders(MAEs)have emerged as a powerful self-supervised approach for point cloud analysis.Nevertheless,existing methods often separately focus on global structures or multi-scale features,ignoring their complementary potential.In this paper,we propose a novel dual-Siamese masked autoencoder(DS-MAE)framework that explores integrating global and hierarchical feature learning in a unified architecture for point cloud analysis.In particular,we introduce a consistent dual-branch patch embedding strategy to partition the point cloud into patches using shared group centers,ensuring both global and hierarchical branches process point patches centered at the same spatial locations.Each branch employs dual-branch Siamese encoders to process original and augmented point patches,learning representations that capture both local details and global context.In addition,we have designed cross-attention Siamese decoders to reconstruct masked point patches and align features both within and between branches with crossattention mechanisms.Comprehensive experiments demonstrate our method consistently achieves superior results to prior methods.Code is available at https://github.com/shaoandy1211/DS-MAE.git.展开更多
Masked autoencoders(MAEs)have recently achieved great success in computer vision.They can automatically extract representations from unlabeled data and improve the performance of various downstream tasks.However,train...Masked autoencoders(MAEs)have recently achieved great success in computer vision.They can automatically extract representations from unlabeled data and improve the performance of various downstream tasks.However,training an MAE model requires substantial resources,which limits their accessibility to many academic institutions:often laboratories in universities lack the necessary resources.This issue significantly hinders the development of this field.In this paper,we propose FastMAE,an efficient MAE approach.Inspired by the idea of offline tokenizers in natural language processing,FastMAE presents a novel way to build an offline vision tokenizer,which can provide high-level semantics in an efficient way.Benefiting from the offline tokenizer,FastMAE becomes an efficient vision learner.Our experiments demonstrate that FastMAE can achieve 83.6%accuracy with ViT-B in only 18.8 h on 8 NVIDIA Tesla-V100 GPUs,which is 31.3×faster than the original MAE,providing a resource friendly baseline for the computer vision community.Moreover,it also achieves comparable performance to state-of-the-art methods.We hope our research will attract more people to engage in MAE-related research and that we can advance its development together.展开更多
Once dwindling,the ancient art of Tibetan Opera is now reaching new stages of development thanks to greater government and audience support Tenzin Yeshe believes he was a Tibetan Opera performer in a previous life.
基金supported by the National Natural Science Foundation of China Funded Project(Project Name:Research on Robust Adaptive Allocation Mechanism of Human Machine Co-Driving System Based on NMS Features,Project Approval Number:52172381).
文摘To address the issue of scarce labeled samples and operational condition variations that degrade the accuracy of fault diagnosis models in variable-condition gearbox fault diagnosis,this paper proposes a semi-supervised masked contrastive learning and domain adaptation(SSMCL-DA)method for gearbox fault diagnosis under variable conditions.Initially,during the unsupervised pre-training phase,a dual signal augmentation strategy is devised,which simultaneously applies random masking in the time domain and random scaling in the frequency domain to unlabeled samples,thereby constructing more challenging positive sample pairs to guide the encoder in learning intrinsic features robust to condition variations.Subsequently,a ConvNeXt-Transformer hybrid architecture is employed,integrating the superior local detail modeling capacity of ConvNeXt with the robust global perception capability of Transformer to enhance feature extraction in complex scenarios.Thereafter,a contrastive learning model is constructed with the optimization objective of maximizing feature similarity across different masked instances of the same sample,enabling the extraction of consistent features from multiple masked perspectives and reducing reliance on labeled data.In the final supervised fine-tuning phase,a multi-scale attention mechanism is incorporated for feature rectification,and a domain adaptation module combining Local Maximum Mean Discrepancy(LMMD)with adversarial learning is proposed.This module embodies a dual mechanism:LMMD facilitates fine-grained class-conditional alignment,compelling features of identical fault classes to converge across varying conditions,while the domain discriminator utilizes adversarial training to guide the feature extractor toward learning domain-invariant features.Working in concert,they markedly diminish feature distribution discrepancies induced by changes in load,rotational speed,and other factors,thereby boosting the model’s adaptability to cross-condition scenarios.Experimental evaluations on the WT planetary gearbox dataset and the Case Western Reserve University(CWRU)bearing dataset demonstrate that the SSMCL-DA model effectively identifies multiple fault classes in gearboxes,with diagnostic performance substantially surpassing that of conventional methods.Under cross-condition scenarios,the model attains fault diagnosis accuracies of 99.21%for the WT planetary gearbox and 99.86%for the bearings,respectively.Furthermore,the model exhibits stable generalization capability in cross-device settings.
文摘In the field of intelligent air combat,real-time and accurate recognition of within-visual-range(WVR)maneuver actions serves as the foundational cornerstone for constructing autonomous decision-making systems.However,existing methods face two major challenges:traditional feature engineering suffers from insufficient effective dimensionality in the feature space due to kinematic coupling,making it difficult to distinguish essential differences between maneuvers,while end-to-end deep learning models lack controllability in implicit feature learning and fail to model high-order long-range temporal dependencies.This paper proposes a trajectory feature pre-extraction method based on a Long-range Masked Autoencoder(LMAE),incorporating three key innovations:(1)Random Fragment High-ratio Masking(RFH-Mask),which enforces the model to learn long-range temporal correlations by masking 80%of trajectory data while retaining continuous fragments;(2)Kalman Filter-Guided Objective Function(KFG-OF),integrating trajectory continuity constraints to align the feature space with kinematic principles;and(3)Two-stage Decoupled Architecture,enabling efficient and controllable feature learning through unsupervised pre-training and frozen-feature transfer.Experimental results demonstrate that LMAE significantly improves the average recognition accuracy for 20-class maneuvers compared to traditional end-to-end models,while significantly accelerating convergence speed.The contributions of this work lie in:introducing high-masking-rate autoencoders into low-informationdensity trajectory analysis,proposing a feature engineering framework with enhanced controllability and efficiency,and providing a novel technical pathway for intelligent air combat decision-making systems.
基金National Natural Science Foundation of China(62472397)Innovation Program for Quantum Science and Technology(2021ZD0302902)。
文摘Recently,several PC oracle based side-channel attacks have been proposed against Kyber.However,most of them focus on unprotected implementations and masking is considered as a counter-measure.In this study,we extend PC oracle based side-channel attacks to the second-order scenario and successfully conduct key-recovery attacks on the first-order masked Kyber.Firstly,we analyze the potential joint information leakage.Inspired by the binary PC oracle based attack proposed by Qin et al.at Asiacrypt 2021,we identify the 1-bit leakage scenario in the masked Keccak implementation.Moreover,we modify the ciphertexts construction described by Tanaka et al.at CHES 2023,extending the leakage scenario from 1-bit to 32-bit.With the assistance of TVLA,we validate these leakages through experiments.Secondly,for these two scenarios,we construct a binary PC oracle based on t-test and a multiple-valued PC oracle based on neural networks.Furthermore,we conduct practical side-channel attacks on masked Kyber by utilizing our oracles,with the implementation running on an ARM Cortex-M4 microcontroller.The demonstrated attacks require a minimum of 15788 and 648 traces to fully recover the key of Kyber768 in the 1-bit leakage scenario and the 32-bit leakage scenario,respectively.Our analysis may also be extended to attack other post-quantum schemes that use the same masked hash function.Finally,we apply the shuffling strategy to the first-order masked imple-mentation of the Kyber and perform leakage tests.Experimental results show that the combination strategy of shuffling and masking can effectively resist our proposed attacks.
基金supported by the project “The demonstration system of rich semantic search application in scientific literature” (Grant No. 1734) from the Chinese Academy of Sciences
文摘Purpose:Mo ve recognition in scientific abstracts is an NLP task of classifying sentences of the abstracts into different types of language units.To improve the performance of move recognition in scientific abstracts,a novel model of move recognition is proposed that outperforms the BERT-based method.Design/methodology/approach:Prevalent models based on BERT for sentence classification often classify sentences without considering the context of the sentences.In this paper,inspired by the BERT masked language model(MLM),we propose a novel model called the masked sentence model that integrates the content and contextual information of the sentences in move recognition.Experiments are conducted on the benchmark dataset PubMed 20K RCT in three steps.Then,we compare our model with HSLN-RNN,BERT-based and SciBERT using the same dataset.Findings:Compared with the BERT-based and SciBERT models,the F1 score of our model outperforms them by 4.96%and 4.34%,respectively,which shows the feasibility and effectiveness of the novel model and the result of our model comes closest to the state-of-theart results of HSLN-RNN at present.Research limitations:The sequential features of move labels are not considered,which might be one of the reasons why HSLN-RNN has better performance.Our model is restricted to dealing with biomedical English literature because we use a dataset from PubMed,which is a typical biomedical database,to fine-tune our model.Practical implications:The proposed model is better and simpler in identifying move structures in scientific abstracts and is worthy of text classification experiments for capturing contextual features of sentences.Originality/value:T he study proposes a masked sentence model based on BERT that considers the contextual features of the sentences in abstracts in a new way.The performance of this classification model is significantly improved by rebuilding the input layer without changing the structure of neural networks.
基金the Project of Introducing Urgently Needed Talents in Key Supporting Regions of Shandong Province,China(No.SDJQP20221805)。
文摘Deep convolutional neural networks(DCNNs)are widely used in content-based image retrieval(CBIR)because of the advantages in image feature extraction.However,the training of deep neural networks requires a large number of labeled data,which limits the application.Self-supervised learning is a more general approach in unlabeled scenarios.A method of fine-tuning feature extraction networks based on masked learning is proposed.Masked autoencoders(MAE)are used in the fine-tune vision transformer(ViT)model.In addition,the scheme of extracting image descriptors is discussed.The encoder of the MAE uses the ViT to extract global features and performs self-supervised fine-tuning by reconstructing masked area pixels.The method works well on category-level image retrieval datasets with marked improvements in instance-level datasets.For the instance-level datasets Oxford5k and Paris6k,the retrieval accuracy of the base model is improved by 7%and 17%compared to that of the original model,respectively.
基金Supported by the National Natural Science Foundation of China（70471057）
文摘We consider a series system of two independent and non-identical components which have different BurrⅫ distributed lifetime.The maximum likelihood and Bayes estimators of the parameters of the system＇s components are obtained based on masked system life test data.The conclusion is that the Bayes estimates are better than the maximum likelihood estimates in the sense of having smaller mean squared errors.
基金supported in part by National Natural Science Foundation of China(No.62176041)in part by Excellent Science and Technique Talent Foundation of Dalian(No.2022RY21).
文摘Significant advancements have beenwitnessed in visual tracking applications leveragingViT in recent years,mainly due to the formidablemodeling capabilities of Vision Transformer(ViT).However,the strong performance of such trackers heavily relies on ViT models pretrained for long periods,limitingmore flexible model designs for tracking tasks.To address this issue,we propose an efficient unsupervised ViT pretraining method for the tracking task based on masked autoencoders,called TrackMAE.During pretraining,we employ two shared-parameter ViTs,serving as the appearance encoder and motion encoder,respectively.The appearance encoder encodes randomly masked image data,while the motion encoder encodes randomly masked pairs of video frames.Subsequently,an appearance decoder and a motion decoder separately reconstruct the original image data and video frame data at the pixel level.In this way,ViT learns to understand both the appearance of images and the motion between video frames simultaneously.Experimental results demonstrate that ViT-Base and ViT-Large models,pretrained with TrackMAE and combined with a simple tracking head,achieve state-of-the-art(SOTA)performance without additional design.Moreover,compared to the currently popular MAE pretraining methods,TrackMAE consumes only 1/5 of the training time,which will facilitate the customization of diverse models for tracking.For instance,we additionally customize a lightweight ViT-XS,which achieves SOTA efficient tracking performance.
基金Supported by the Postgraduate Research and Practice Innovation Program of Nanjing University of Aeronautics and Astronautics(XCXJH20220318)。
文摘Since the outbreak of Coronavirus Disease 2019(COVID-19),people are recommended to wear facial masks to limit the spread of the virus.Under the circumstances,traditional face recognition technologies cannot achieve satisfactory results.In this paper,we propose a face recognition algorithm that combines the traditional features and deep features of masked faces.For traditional features,we extract Local Binary Pattern(LBP),Scale-Invariant Feature Transform(SIFT)and Histogram of Oriented Gradient(HOG)features from the periocular region,and use the Support Vector Machines(SVM)classifier to perform personal identification.We also propose an improved Convolutional Neural Network(CNN)model Angular Visual Geometry Group Network(A-VGG)to learn deep features.Then we use the decision-level fusion to combine the four features.Comprehensive experiments were carried out on databases of real masked faces and simulated masked faces,including frontal and side faces taken at different angles.Images with motion blur were also tested to evaluate the robustness of the algorithm.Besides,the experiment of matching a masked face with the corresponding full face is accomplished.The experimental results show that the proposed algorithm has state-of-the-art performance in masked face recognition,and the periocular region has rich biological features and high discrimination.
基金supported by the National Natural Science Foundation of China(71401134 71571144+1 种基金 71171164)the Program of International Cooperation and Exchanges in Science and Technology Funded by Shaanxi Province(2016KW-033)
文摘Under Type-Ⅱ progressively hybrid censoring, this paper discusses statistical inference and optimal design on stepstress partially accelerated life test for hybrid system in presence of masked data. It is assumed that the lifetime of the component in hybrid systems follows independent and identical modified Weibull distributions. The maximum likelihood estimations(MLEs)of the unknown parameters, acceleration factor and reliability indexes are derived by using the Newton-Raphson algorithm. The asymptotic variance-covariance matrix and the approximate confidence intervals are obtained based on normal approximation to the asymptotic distribution of MLEs of model parameters. Moreover,two bootstrap confidence intervals are constructed by using the parametric bootstrap method. The optimal time of changing stress levels is determined under D-optimality and A-optimality criteria.Finally, the Monte Carlo simulation study is carried out to illustrate the proposed procedures.
文摘The fat production rate in adult healthy masked civet(Paguma lavata) and nutria (Myocaster coypus) oil were measured. The values of iodine. saponification and acid PH, composition of fstty acids of grease were analyzed both chemically and by apparatus. The results showed that acid PH, iodine value, saponification value,and unsaturation point are 1.887 and 0.784, 53.90 and 48.32, 98.80 and 100.23. and 60.05% and 58.85% are respectively for masked civet's fat and nutria's oil. Both of masked civet's fat and nutria's oil contain a little of Eicosatetraenoic acid (C-20;4), which is of great significance in nutrition and metabolism for human body. The analysis results indicate that masked civet's oil is similar to nutria's oil in iodine value, saponification value and unsaturation point. Both masked civet's fat and nutria's oil are steady and have highly nutrition. They can be widely exploited and utilized in health protection and cosmetics made industry.
文摘Background Hypertension is the main risk factor for cardiovascular diseases, affecting more than half the elderly population. It is essential to know if they have proper control of hypertension. The aim of this study was to identify the associated factors to masked uncon- trolled hypertension and false uncontrolled hypertension in older patients. Methods Two-hundred seventy-three individuals （70.1±6.7 years-old） had blood pressure （BP） measured at the office and by ambulatory BP monitoring （ABPM）, with the definition of controlled group （C）, individuals with high office BP and adequate ABPM, called white-coat effect group （WCE）, uncontrolled （UC）, and subjects with ap- propriate office BP and elevated ABPM denominated masked effect group （ME）. Age, body mass index, diabetes, pulse pressure （PP） and BP dipping during sleep were evaluated （Kruskal-Wallis test and logistic regression models）. Results Age was higher in UC than in C and ME （P 〈 0.01）, and 24-h ABPM PP was lower in C （48± 7 mmHg） and WCE （51±6 mmHg） than in UC （67±12 mmHg） and ME （59±8 mmHg） （P 〈 0.01）. Sleep systolic BP dipping was lower in ME than in C （P = 0.03）. Female gender was associated with a greater chance of being of ME group, which showed a higher PP and lower BP dipping during sleep. Conclusions In older individuals, office BP measure- ments did not allow the detection of associated factors that would permit to differentiate WCE from UC group and C from ME group. ABPM favored the identification of a higher PP and a lower BP dipping during sleep in the masked effect and uncontrolled groups.
文摘Objective:To compare the risk of target organ damage in masked hypertension(MH)and sustained hypertension(SH).Methods:A systematic review and meta-analysis was performed.A search of PubMed,Embase,and the Cochrane Library of relevant case-control studies was performed from inception to December 2019,and articles on MH and SH selected according to the inclusion criteria were analyzed.The primary end point was target organ damage in the heart.The secondary end points were target organ damage in the kidneys and blood vessels.Results:Seventeen studies that met the screening criteria were included in the meta-analysis.Compared with the SH group,in the MH group carotid intima-media thickness(IMT)and E/A ratio were signifi cantly greater and the prevalence of left ventricular remodeling and the pulse wave velocity were signifi cantly lower.Other indicators in the heart,kidneys,and blood vessels were not statistically different between the two groups.IMT:P=0.01,E/A ratio:P=0.01,prevalence of left ventricular remodeling:P=0.02,pulse wave velocity:P=0.01.Conclusion:Our study has shown that MH may have almost the same degree of target organ damage as SH,so clinicians may need to consider target organ damage.
文摘The purpose of this study was to cover the bitter taste of arbidol hydrochloride(ARB)and develop dry suspension with combination of solid dispersion and flavors.Taste masking was successfully done by solid dispersion using octadecanol as the carrier by fusion method.Suspending agents,carriers and other excipients were selected.Differential scanning calorimetry(DSC)and Fourier transform infrared spectroscopy(FTIR)were performed to identify the physicochemical interaction between drug and carrier,DSC analysis indicated that ARB was amorphous in the solid dispersion,FTIR spectroscopy showed no interaction between drug and carrier.Taste masking was evaluated on six volunteers with a score of 4.9.The results demonstrated successful taste masking.Water was used to study the in vitro dissolution performance of the three formulations of commercial tablet,capsule and self-made suspension.The self-made suspension showed a lower and slower release,the insoluble carrier octadecanol blocked the drug dissolving from the solid dispersion.It was indicated from the primary stability study,the self-made suspensions were sensitive to high temperature,high humidity and strong light conditions,they should be stored in sealed containers away from heat,light and humidity.
基金the National Nature Science Foundation of China(No.81173008)from Project for Excellent Talents of Liaoning Province(No.LR20110028)from Program for New Century Excellent Talents in University(No.NCET-12-1015).
文摘Cefuroxime axetil(CA)is an ester prodrug of cefuroxime with an unpleasant taste when administrated orally.This work was to mask the bitter taste of CA and enhance its oral bioavailability.Dry suspensions were prepared by means of wet granulation method and solid dispersion method.Binders,suspending agents and other compositions involved in the formulation were optimized.The differential scanning calorimetry(DSC)analysis indicated that CA was amorphous in the solid dispersion with stearic acid as the carrier,which contributed to an improvement of the dissolution rate.Taste evaluation was performed by three volunteers and taste masking was successfully achieved by the methods mentioned above.A pH 7.0 phosphate buffer was adopted to study the in vitro dissolution performance of the three formulations,i.e.,two self-made dry suspensions and the commercial one.With a better release characteristic and a satisfying taste masking ability,the solid dispersion suspension was selected as the optimal formulation for the further pharmacokinetic study in beagle dogs.The values of Cmax and AUC0e12 for the solid dispersion suspension were about 1.78-fold and 2.17-fold higher than these of reference suspension,respectively.The obtained results demonstrated that the solid dispersion can efficiently mask the bitter taste of CA and significantly enhance its oral bioavailability.
文摘Civets are small mammals belonging to the family Viverridae.The masked palm civets(Paguma larvata)served as an intermediate host in the bat-to-human transmission of severe acute respiratory syndrome coronavirus(SARS-Co V)in 2003(Guan et al.,2003).Because of their unique role in the SARS outbreak,civets were suspected as a potential intermediate host of SARS-Co V-2.
基金Princess Nourah bint Abdulrahman University Researchers Supporting Project Number (PNURSP2023R442),Princess Nourah bint Abdulrahman University,Riyadh,Saudi Arabia。
文摘Face mask detection has several applications,including real-time surveillance,biometrics,etc.Identifying face masks is also helpful for crowd control and ensuring people wear them publicly.With monitoring personnel,it is impossible to ensure that people wear face masks;automated systems are a much superior option for face mask detection and monitoring.This paper introduces a simple and efficient approach for masked face detection.The architecture of the proposed approach is very straightforward;it combines deep learning and local binary patterns to extract features and classify themasmasked or unmasked.The proposed systemrequires hardware withminimal power consumption compared to state-of-the-art deep learning algorithms.Our proposed system maintains two steps.At first,this work extracted the local features of an image by using a local binary pattern descriptor,and then we used deep learning to extract global features.The proposed approach has achieved excellent accuracy and high performance.The performance of the proposed method was tested on three benchmark datasets:the realworld masked faces dataset(RMFD),the simulated masked faces dataset(SMFD),and labeled faces in the wild(LFW).Performancemetrics for the proposed technique weremeasured in terms of accuracy,precision,recall,and F1-score.Results indicated the efficiency of the proposed technique,providing accuracies of 99.86%,99.98%,and 100%for RMFD,SMFD,and LFW,respectively.Moreover,the proposed method outperformed state-of-the-art deep learning methods in the recent bibliography for the same problem under study and on the same evaluation datasets.
文摘Masked autoencoders(MAEs)have emerged as a powerful self-supervised approach for point cloud analysis.Nevertheless,existing methods often separately focus on global structures or multi-scale features,ignoring their complementary potential.In this paper,we propose a novel dual-Siamese masked autoencoder(DS-MAE)framework that explores integrating global and hierarchical feature learning in a unified architecture for point cloud analysis.In particular,we introduce a consistent dual-branch patch embedding strategy to partition the point cloud into patches using shared group centers,ensuring both global and hierarchical branches process point patches centered at the same spatial locations.Each branch employs dual-branch Siamese encoders to process original and augmented point patches,learning representations that capture both local details and global context.In addition,we have designed cross-attention Siamese decoders to reconstruct masked point patches and align features both within and between branches with crossattention mechanisms.Comprehensive experiments demonstrate our method consistently achieves superior results to prior methods.Code is available at https://github.com/shaoandy1211/DS-MAE.git.
基金supported by the National Science and Technology Major Project(Grant No.2021ZD0112902)the National Natural Science Foundation of China(Grant Nos.623B2057 and 62220106003)+1 种基金Tsinghua University Initiative Scientific Research ProgramTsinghua-Tencent Joint Laboratory for Internet Innovation Technology.
文摘Masked autoencoders(MAEs)have recently achieved great success in computer vision.They can automatically extract representations from unlabeled data and improve the performance of various downstream tasks.However,training an MAE model requires substantial resources,which limits their accessibility to many academic institutions:often laboratories in universities lack the necessary resources.This issue significantly hinders the development of this field.In this paper,we propose FastMAE,an efficient MAE approach.Inspired by the idea of offline tokenizers in natural language processing,FastMAE presents a novel way to build an offline vision tokenizer,which can provide high-level semantics in an efficient way.Benefiting from the offline tokenizer,FastMAE becomes an efficient vision learner.Our experiments demonstrate that FastMAE can achieve 83.6%accuracy with ViT-B in only 18.8 h on 8 NVIDIA Tesla-V100 GPUs,which is 31.3×faster than the original MAE,providing a resource friendly baseline for the computer vision community.Moreover,it also achieves comparable performance to state-of-the-art methods.We hope our research will attract more people to engage in MAE-related research and that we can advance its development together.
文摘Once dwindling,the ancient art of Tibetan Opera is now reaching new stages of development thanks to greater government and audience support Tenzin Yeshe believes he was a Tibetan Opera performer in a previous life.
