Computed Tomography(CT)reconstruction is essential inmedical imaging and other engineering fields.However,blurring of the projection during CT imaging can lead to artifacts in the reconstructed images.Projection blur ...Computed Tomography(CT)reconstruction is essential inmedical imaging and other engineering fields.However,blurring of the projection during CT imaging can lead to artifacts in the reconstructed images.Projection blur combines factors such as larger ray sources,scattering and imaging system vibration.To address the problem,we propose DeblurTomo,a novel self-supervised learning-based deblurring and reconstruction algorithm that efficiently reconstructs sharp CT images from blurry input without needing external data and blur measurement.Specifically,we constructed a coordinate-based implicit neural representation reconstruction network,which can map the coordinates to the attenuation coefficient in the reconstructed space formore convenient ray representation.Then,wemodel the blur as aweighted sumof offset rays and design the RayCorrectionNetwork(RCN)andWeight ProposalNetwork(WPN)to fit these rays and their weights bymulti-view consistency and geometric information,thereby extending 2D deblurring to 3D space.In the training phase,we use the blurry input as the supervision signal to optimize the reconstruction network,the RCN,and the WPN simultaneously.Extensive experiments on the widely used synthetic dataset show that DeblurTomo performs superiorly on the limited-angle and sparse-view in the simulated blurred scenarios.Further experiments on real datasets demonstrate the superiority of our method in practical scenarios.展开更多
The isolation of healthcare data among worldwide hospitals and institutes forms barriers for fully realizing the data-hungry artificial intelligence(AI)models promises in renewing medical services.To overcome this,pri...The isolation of healthcare data among worldwide hospitals and institutes forms barriers for fully realizing the data-hungry artificial intelligence(AI)models promises in renewing medical services.To overcome this,privacy-preserving distributed learning frameworks,represented by swarm learning and federated learning,have been investigated recently with the sensitive healthcare data retaining in its local premises.However,existing frameworks use a one-size-fits-all mode that tunes one model for all healthcare situations,which could hardly fit the usually diverse disease prediction in practice.This work introduces the idea of ensemble learning into privacypreserving distributed learning and presents the En-split framework,where the predictions of multiple expert models with specialized diagnostic capabilities are jointly explored.Considering the exacerbation of communication and computation burdens with multiple models during learning,model split is used to partition targeted models into two parts,with hospitals focusing on building the feature-enriched shallow layers.Meanwhile,dedicated noises are implemented to the edge layers for differential privacy protection.Experiments on two public datasets demonstrate En-split’s superior performance on accuracy and efficiency,compared with existing distributed learning frameworks.展开更多
BACKGROUND： Evidence illustrates that androgen has a neuroprotective role. However, whether androgen also has the protective effect on hippocampal neurons during free radical mediated injury remains unclear. OBJECTIV...BACKGROUND： Evidence illustrates that androgen has a neuroprotective role. However, whether androgen also has the protective effect on hippocampal neurons during free radical mediated injury remains unclear. OBJECTIVE： To investigate the neuroprotective effect of androgen on hippocampal neurons during free radical damage. DESIGN, TIME AND SETTING： A controlled in vitro experiment was performed at the Department of Human Anatomy, Cell Culture Lab, and Neuroendocrinology Lab, Basic Medical School, Hebei Medical University from February to June 2009. MATERIALS： Testosterone was provided by Tianjin Jinyao Amino Acid Company, China. METHODS： Primary cultured neurons from 24 Sprague Dawley rats were randomly assigned into four groups： control, H202, testosterone, and testosterone （pre-added） plus H2O2 groups. MAIN OUTCOME MEASURES： The positive cell ratio of microtubule associated protein-Ⅱ and neuron specific enolase was determined by immunocytochemistry. Neuronal morphology was observed by hematoxylin-eosin staining and Nissl staining. Cell vitality and viability were determined using an inverted phase contrast microscope. The content of nitric oxide synthase, malondialdehyde, and superoxide dismutase were measured with a spectrophotometer. RESULTS： As compared with the control group, cell vitality and viability, and superoxide dismutase level were significantly decreased in the H202 group （P 〈 0.05）, while nitric oxide synthase and malondialdehyde levels were significantly increased （P 〈 0.05）. Neuronal vitality and viability as well as superoxide dismutase level in the testosterone plus H2O2 group were significantly greater than in the H2O2 group （P 〈 0.05）, and nitric oxide synthase and malondialdehyde levels were significantly less than in the H2O2 group （P〈 0.05）. CONCLUSION： Androgen partially reversed H2O2-induced neuronal damage and protected neurons.展开更多
Here we developed a novel wavelength-switchable visible continuous-wave(CW)Pr^(3+):YLF laser around 670 nm.In single-wavelength laser operations,the maximum output powers of 2.60 W,1.26 W,and 0.21 W,the maximum slope ...Here we developed a novel wavelength-switchable visible continuous-wave(CW)Pr^(3+):YLF laser around 670 nm.In single-wavelength laser operations,the maximum output powers of 2.60 W,1.26 W,and 0.21 W,the maximum slope efficiencies of 34.7%,27.3%,and 12.3%were achieved with good beam qualities(M^(2)<1.6)at 670.4 nm,674.2 nm,and 678.9 nm,respectively.Record-high output power(2.6 W)and record-high slope efficiency(34.7%)were achieved for the Pr^(3+):YLF laser operation at 670.4 nm.This is also the first demonstration of longer-wavelength peaks beyond 670 nm in the^(3)P_(1)→^(3)F_(3)transition of Pr^(3+):YLF.In multi-wavelength laser operations,the dual-wavelength lasings,including 670.1/674.8 nm,670.1/679.1 nm,and 675.0/679.4 nm,were obtained by fine adjustment of one/two etalons within the cavity.Furthermore,the triple-wavelength lasings,e.g.672.2/674.2/678.6 nm and 670.4/674.8/679.4 nm,were successfully demonstrated.Moreover,both the first-order vortex lasers(LG_(0)^(+1)and LG_(0)^(-1)modes)at 670.4 nm were obtained by off-axis pumping.展开更多
The extreme imbalanced data problem is the core issue in anomaly detection.The amount of abnormal data is so small that we cannot get adequate information to analyze it.The mainstream methods focus on taking fully adv...The extreme imbalanced data problem is the core issue in anomaly detection.The amount of abnormal data is so small that we cannot get adequate information to analyze it.The mainstream methods focus on taking fully advantages of the normal data,of which the discrimination method is that the data not belonging to normal data distribution is the anomaly.From the view of data science,we concentrate on the abnormal data and generate artificial abnormal samples by machine learning method.In this kind of technologies,Synthetic Minority Over-sampling Technique and its improved algorithms are representative milestones,which generate synthetic examples randomly in selected line segments.In our work,we break the limitation of line segment and propose an Imbalanced Triangle Synthetic Data method.In theory,our method covers a wider range.In experiment with real world data,our method performs better than the SMOTE and its meliorations.展开更多
By using efficient and timely medical diagnostic decision making,clinicians can positively impact the quality and cost of medical care.However,the high similarity of clinical manifestations between diseases and the li...By using efficient and timely medical diagnostic decision making,clinicians can positively impact the quality and cost of medical care.However,the high similarity of clinical manifestations between diseases and the limitation of clinicians’knowledge both bring much difficulty to decision making in diagnosis.Therefore,building a decision support system that can assist medical staff in diagnosing and treating diseases has lately received growing attentions in the medical domain.In this paper,we employ a multi-label classification framework to classify the Chinese electronic medical records to establish corresponding relation between the medical records and disease categories,and compare this method with the traditional medical expert system to verify the performance.To select the best subset of patient features,we propose a feature selection method based on the composition and distribution of symptoms in electronic medical records and compare it with the traditional feature selection methods such as chi-square test.We evaluate the feature selection methods and diagnostic models from two aspects,false negative rate(FNR)and accuracy.Extensive experiments have conducted on a real-world Chinese electronic medical record database.The evaluation results demonstrate that our proposed feature selection method can improve the accuracy and reduce the FNR compare to the traditional feature selection methods,and the multi-label classification framework have better accuracy and lower FNR than the traditional expert system.展开更多
With the continuous development of e-commerce,consumers show increasing interest in posting comments on consumption experience and quality of commodities.Meanwhile,people make purchasing decisions relying on other com...With the continuous development of e-commerce,consumers show increasing interest in posting comments on consumption experience and quality of commodities.Meanwhile,people make purchasing decisions relying on other comments much more than ever before.So the reliability of commodity comments has a significant impact on ensuring consumers’equity and building a fair internet-trade-environment.However,some unscrupulous online-sellers write fake praiseful reviews for themselves and malicious comments for their business counterparts to maximize their profits.Those improper ways of self-profiting have severely ruined the entire online shopping industry.Aiming to detect and prevent these deceptive comments effectively,we construct a model of Multi-Filters Convolutional Neural Network(MFCNN)for opinion spam detection.MFCNN is designed with a fixed-length sequence input and an improved activation function to avoid the gradient vanishing problem in spam opinion detection.Moreover,convolution filters with different widths are used in MFCNN to represent the sentences and documents.Our experimental results show that MFCNN outperforms current state-of-the-art methods on standard spam detection benchmarks.展开更多
Mid-infrared(MIR)fiber pulsed lasers are of tremendous application interest in eye-safe LIDAR,spectroscopy,chemical detection and medicine.So far,these MIR lasers largely required bulk optical elements,complex free-sp...Mid-infrared(MIR)fiber pulsed lasers are of tremendous application interest in eye-safe LIDAR,spectroscopy,chemical detection and medicine.So far,these MIR lasers largely required bulk optical elements,complex free-space light alignment and large footprint,precluding compact all-fiber structure.Here,we proposed and demonstrated an all-fiberized structured gain-switched Ho3+-doped ZBLAN fiber laser operating around 2.9μm.A home-made 1146 nm Raman fiber pulsed laser was utilized to pump highly concentrated single-cladding Ho3+-doped ZBLAN fiber with different lengths of 2 m or 0.25 m.A home-made MIR fiber mirror and a perpendicular-polished ZBLAN fiber end construct the all-fiberized MIR cavity.Stable gain-switched multiple states with a sub-pulse number tuned from 1 to 8 were observed.The effects of gain fiber length,pump power,pump repetition rate and output coupling ratio on performance of gain-switched pulses were further investigated in detail.The shortest pulse duration of 283 ns was attained with 10 kHz repetition rate.The pulsed laser,centered at 2.92μm,had a maximum average output power of 54.2 mW and a slope efficiency of 10.12%.It is,to the best of our knowledge,the first time to demonstrate a mid-infrared gain-switched Ho3+:ZBLAN fiber laser with compact all-fiber structure.展开更多
Malicious attacks can be launched by misusing the network address translation technique as a camouflage.To mitigate such threats,network address translation identification is investigated to identify network address t...Malicious attacks can be launched by misusing the network address translation technique as a camouflage.To mitigate such threats,network address translation identification is investigated to identify network address translation devices and detect abnormal behaviors.However,existingmethods in this field are mainly developed for relatively small-scale networks and work in an offline manner,which cannot adapt to the real-time inference requirements in high-speed network scenarios.In this paper,we propose a flexible and efficient network address translation identification scheme based on actively measuring the distance of a round trip to a target with decremental time-tolive values.The basic intuition is that the incoming and outgoing traffic froma network address translation device usually experiences the different number of hops,which can be discovered by probing with dedicated time-to-live values.We explore a joint effort of parallel transmission,stateless probes,and flexible measuring reuse to accommodate the efficiency of the measuring process.We further accelerate statistical countingwith a new sublinear space data structure Bi-sketch.We implement a prototype and conduct real-world deployments with 1000 volunteers in 31 Chinese provinces,which is believed to bring insight for ground truth collection in this field.Experiments onmulti-sources datasets show that our proposal can achieve as high precision and recall as 95%with a traffic handling throughput of over 106 pps.展开更多
Alias resolution,mapping IP addresses to routers,is a critical step in obtaining a network topology.The latest work on alias resolution is based on special fields in the packet,such as IP ID,port number,etc.However,fo...Alias resolution,mapping IP addresses to routers,is a critical step in obtaining a network topology.The latest work on alias resolution is based on special fields in the packet,such as IP ID,port number,etc.However,for security reasons,most network devices block packets for setting options,and some related fields exist only in IPv4,so these methods cannot be used for alias resolution of IPv6.In order to solve the above problems,we propose an alias analysis method based on delay sequence analysis.In this article,we present a new model to describe the distribution of Internet delays and give a mathematical proof.After experimental measurements using the Macroscopic Internet Topology Data Kit(ITDK)and Ark IPv6 Topology Dataset,it was found that the statistical differences in most alias delay models were very small.The statistical differences in the non-alias delay models are spread over a wide range.Using the wavelet decomposition in delay sequence,it was found that the approximate components and the detail components of the delay sequence of aliases were the same after filtering out the noise,which provided a theoretical explanation for the experimental results.This technology is applicable to both IPv4 and IPv6.展开更多
For many Internet companies,a huge amount of KPIs(e.g.,server CPU usage,network usage,business monitoring data)will be generated every day.How to closely monitor various KPIs,and then quickly and accurately detect ano...For many Internet companies,a huge amount of KPIs(e.g.,server CPU usage,network usage,business monitoring data)will be generated every day.How to closely monitor various KPIs,and then quickly and accurately detect anomalies in such huge data for troubleshooting and recovering business is a great challenge,especially for unlabeled data.The generated KPIs can be detected by supervised learning with labeled data,but the current problem is that most KPIs are unlabeled.That is a time-consuming and laborious work to label anomaly for company engineers.Build an unsupervised model to detect unlabeled data is an urgent need at present.In this paper,unsupervised learning DBSCAN combined with feature extraction of data has been used,and for some KPIs,its best F-Score can reach about 0.9,which is quite good for solving the current problem.展开更多
The research and analysis of Internet topology is hot in the field of networkmeasurement, which have important applications in network security, traffic schedulingand many other fields. Most of the existing works are ...The research and analysis of Internet topology is hot in the field of networkmeasurement, which have important applications in network security, traffic schedulingand many other fields. Most of the existing works are focused on the AS-level and routerleveltopology, but few works are about the IP-level topology. In fact, obtaining thetopology of each continent and knowing how the topologies of the continents areconnected to each other can help us understanding the Internet around the world morethoroughly. In this paper, we obtained data sets from RIPE, constructed and analyzednetwork topologies of all the continents. By analyzing the topological connectionsbetween continents, we found out that most of the junctions of inter-continent traces arelocated in a few countries.展开更多
While smart devices based on ARM processor bring us a lot of convenience,they also become an attractive target of cyber-attacks.The threat is exaggerated as commodity OSes usually have a large code base and suffer fro...While smart devices based on ARM processor bring us a lot of convenience,they also become an attractive target of cyber-attacks.The threat is exaggerated as commodity OSes usually have a large code base and suffer from various software vulnerabilities.Nowadays,adversaries prefer to steal sensitive data by leaking the content of display output by a security-sensitive application.A promising solution is to exploit the hardware visualization extensions provided by modern ARM processors to construct a secure display path between the applications and the display device.In this work,we present a scheme named SecDisplay for trusted display service,it protects sensitive data displayed from being stolen or tampered surreptitiously by a compromised OS.The TCB of SecDisplay mainly consists of a tiny hypervisor and a super light-weight rendering painter,and has only^1400 lines of code.We implemented a prototype of SecDisplay and evaluated its performance overhead.The results show that SecDisplay only incurs an average drop of 3.4%.展开更多
The global Internet is a complex network of interconnected autonomous systems(ASes).Understanding Internet inter-domain path information is crucial for understanding,managing,and improving the Internet.The path inform...The global Internet is a complex network of interconnected autonomous systems(ASes).Understanding Internet inter-domain path information is crucial for understanding,managing,and improving the Internet.The path information can also help protect user privacy and security.However,due to the complicated and heterogeneous structure of the Internet,path information is not publicly available.Obtaining path information is challenging due to the limited measurement probes and collectors.Therefore,inferring Internet inter-domain paths from the limited data is a supplementary approach to measure Internet inter-domain paths.The purpose of this survey is to provide an overview of techniques that have been conducted to infer Internet inter-domain paths from 2005 to 2023 and present the main lessons from these studies.To this end,we summarize the inter-domain path inference techniques based on the granularity of the paths,for each method,we describe the data sources,the key ideas,the advantages,and the limitations.To help readers understand the path inference techniques,we also summarize the background techniques for path inference,such as techniques to measure the Internet,infer AS relationships,resolve aliases,and map IP addresses to ASes.A case study of the existing techniques is also presented to show the real-world applications of inter-domain path inference.Additionally,we discuss the challenges and opportunities in inferring Internet inter-domain paths,the drawbacks of the state-of-the-art techniques,and the future directions.展开更多
Time series segmentation has attracted more interests in recent years,which aims to segment time series into different segments,each reflects a state of the monitored objects.Although there have been many surveys on t...Time series segmentation has attracted more interests in recent years,which aims to segment time series into different segments,each reflects a state of the monitored objects.Although there have been many surveys on time series segmentation,most of them focus more on change point detection(CPD)methods and overlook the advances in boundary detection(BD)and state detection(SD)methods.In this paper,we categorize time series segmentation methods into CPD,BD,and SD methods,with a specific focus on recent advances in BD and SD methods.Within the scope of BD and SD,we subdivide the methods based on their underlying models/techniques and focus on the milestones that have shaped the development trajectory of each category.As a conclusion,we found that:(1)Existing methods failed to provide sufficient support for online working,with only a few methods supporting online deployment;(2)Most existing methods require the specification of parameters,which hinders their ability to work adaptively;(3)Existing SD methods do not attach importance to accurate detection of boundary points in evaluation,which may lead to limitations in boundary point detection.We highlight the ability to working online and adaptively as important attributes of segmentation methods,the boundary detection accuracy as a neglected metrics for SD methods.展开更多
As an emerging discipline,machine learning has been widely used in artificial intelligence,education,meteorology and other fields.In the training of machine learning models,trainers need to use a large amount of pract...As an emerging discipline,machine learning has been widely used in artificial intelligence,education,meteorology and other fields.In the training of machine learning models,trainers need to use a large amount of practical data,which inevitably involves user privacy.Besides,by polluting the training data,a malicious adversary can poison the model,thus compromising model security.The data provider hopes that the model trainer can prove to them the confidentiality of the model.Trainer will be required to withdraw data when the trust collapses.In the meantime,trainers hope to forget the injected data to regain security when finding crafted poisoned data after the model training.Therefore,we focus on forgetting systems,the process of which we call machine unlearning,capable of forgetting specific data entirely and efficiently.In this paper,we present the first comprehensive survey of this realm.We summarize and categorize existing machine unlearning methods based on their characteristics and analyze the relation between machine unlearning and relevant fields(e.g.,inference attacks and data poisoning attacks).Finally,we briefly conclude the existing research directions.展开更多
Deep learning technology has been widely used in computer vision,speech recognition,natural language processing,and other related fields.The deep learning algorithm has high precision and high reliability.However,the ...Deep learning technology has been widely used in computer vision,speech recognition,natural language processing,and other related fields.The deep learning algorithm has high precision and high reliability.However,the lack of resources in the edge terminal equipment makes it difficult to run deep learning algorithms that require more memory and computing power.In this paper,we propose MoTransFrame,a general model processing framework for deep learning models.Instead of designing a model compression algorithm with a high compression ratio,MoTransFrame can transplant popular convolutional neural networks models to resources-starved edge devices promptly and accurately.By the integration method,Deep learning models can be converted into portable projects for Arduino,a typical edge device with limited resources.Our experiments show that MoTransFrame has good adaptability in edge devices with limited memories.It is more flexible than other model transplantation methods.It can keep a small loss of model accuracy when the number of parameters is compressed by tens of times.At the same time,the computational resources needed in the reasoning process are less than what the edge node could handle.展开更多
In the era of internet proliferation,safeguarding digital media copyright and integrity,especially for images,is imperative.Digital watermarking stands out as a pivotal solution for image security.With the advent of d...In the era of internet proliferation,safeguarding digital media copyright and integrity,especially for images,is imperative.Digital watermarking stands out as a pivotal solution for image security.With the advent of deep learning,watermarking has seen significant advancements.Our review focuses on the innovative deep watermarking approaches that employ neural networks to identify robust embedding spaces,resilient to various attacks.These methods,characterized by a streamlined encoder-decoder architecture,have shown enhanced performance through the incorporation of novel training modules.This article offers an in-depth analysis of deep watermarking’s core technologies,current status,and prospective trajectories,evaluating recent scholarly contributions across diverse frameworks.It concludes with an overview of the technical hurdles and prospects,providing essential insights for ongoing and future research endeavors in digital image watermarking.展开更多
Wearable devices are becoming more popular in our daily life.They are usually used to monitor health status,track fitness data,or even do medical tests,etc.Since the wearable devices can obtain a lot of personal data,...Wearable devices are becoming more popular in our daily life.They are usually used to monitor health status,track fitness data,or even do medical tests,etc.Since the wearable devices can obtain a lot of personal data,their security issues are very important.Motivated by the consideration that the current pairing mechanisms of Bluetooth Low Energy(BLE)are commonly impractical or insecure for many BLE based wearable devices nowadays,we design and implement a security framework in order to protect the communication between these devices.The security framework is a supplement to the Bluetooth pairing mechanisms and is compatible with all BLE based wearable devices.The framework is a module between the application layer and the GATT(Generic Attribute Profile)layer in the BLE architecture stack.When the framework starts,a client and a server can automatically and securely establish shared fresh keys following a designed protocol;the services of encrypting and decrypting messages are provided to the applications conveniently by two functions;application data are securely transmitted following another protocol using the generated keys.Prudential principles are followed by the design of the framework for security purposes.It can protect BLE based wearable devices from replay attacks,Man-in-The-Middle attacks,data tampering,and passive eavesdropping.We conduct experiments to show that the framework can be conveniently deployed with practical operational cost of power consumption.The protocols in this framework have been formally verified that the designed security goals are satisfied.展开更多
In geometry processing,symmetry research benefits from global geo-metric features of complete shapes,but the shape of an object captured in real-world applications is often incomplete due to the limited sensor resoluti...In geometry processing,symmetry research benefits from global geo-metric features of complete shapes,but the shape of an object captured in real-world applications is often incomplete due to the limited sensor resolution,single viewpoint,and occlusion.Different from the existing works predicting symmetry from the complete shape,we propose a learning approach for symmetry predic-tion based on a single RGB-D image.Instead of directly predicting the symmetry from incomplete shapes,our method consists of two modules,i.e.,the multi-mod-al feature fusion module and the detection-by-reconstruction module.Firstly,we build a channel-transformer network(CTN)to extract cross-fusion features from the RGB-D as the multi-modal feature fusion module,which helps us aggregate features from the color and the depth separately.Then,our self-reconstruction net-work based on a 3D variational auto-encoder(3D-VAE)takes the global geo-metric features as input,followed by a prediction symmetry network to detect the symmetry.Our experiments are conducted on three public datasets:ShapeNet,YCB,and ScanNet,we demonstrate that our method can produce reliable and accurate results.展开更多
基金supported in part by the National Natural Science Foundation of China under Grants 62472434 and 62402171in part by the National Key Research and Development Program of China under Grant 2022YFF1203001+1 种基金in part by the Science and Technology Innovation Program of Hunan Province under Grant 2022RC3061in part by the Sci-Tech Innovation 2030 Agenda under Grant 2023ZD0508600.
文摘Computed Tomography(CT)reconstruction is essential inmedical imaging and other engineering fields.However,blurring of the projection during CT imaging can lead to artifacts in the reconstructed images.Projection blur combines factors such as larger ray sources,scattering and imaging system vibration.To address the problem,we propose DeblurTomo,a novel self-supervised learning-based deblurring and reconstruction algorithm that efficiently reconstructs sharp CT images from blurry input without needing external data and blur measurement.Specifically,we constructed a coordinate-based implicit neural representation reconstruction network,which can map the coordinates to the attenuation coefficient in the reconstructed space formore convenient ray representation.Then,wemodel the blur as aweighted sumof offset rays and design the RayCorrectionNetwork(RCN)andWeight ProposalNetwork(WPN)to fit these rays and their weights bymulti-view consistency and geometric information,thereby extending 2D deblurring to 3D space.In the training phase,we use the blurry input as the supervision signal to optimize the reconstruction network,the RCN,and the WPN simultaneously.Extensive experiments on the widely used synthetic dataset show that DeblurTomo performs superiorly on the limited-angle and sparse-view in the simulated blurred scenarios.Further experiments on real datasets demonstrate the superiority of our method in practical scenarios.
基金supported by the National Natural Science Foundation of China(62172155)the NationalKey Research andDevelopment Programof China(2022YFF1203001)+2 种基金the Science and Technology Innovation Program of Hunan Province(Nos.2022RC3061,2023RC3027)the Graduate Research Innovation Project of Hunan Province(XJCX2023157)NUDT Scientific Project“Research on Privacy-Enhancing Computing Technologies for Activity Trajectory Data”.
文摘The isolation of healthcare data among worldwide hospitals and institutes forms barriers for fully realizing the data-hungry artificial intelligence(AI)models promises in renewing medical services.To overcome this,privacy-preserving distributed learning frameworks,represented by swarm learning and federated learning,have been investigated recently with the sensitive healthcare data retaining in its local premises.However,existing frameworks use a one-size-fits-all mode that tunes one model for all healthcare situations,which could hardly fit the usually diverse disease prediction in practice.This work introduces the idea of ensemble learning into privacypreserving distributed learning and presents the En-split framework,where the predictions of multiple expert models with specialized diagnostic capabilities are jointly explored.Considering the exacerbation of communication and computation burdens with multiple models during learning,model split is used to partition targeted models into two parts,with hospitals focusing on building the feature-enriched shallow layers.Meanwhile,dedicated noises are implemented to the edge layers for differential privacy protection.Experiments on two public datasets demonstrate En-split’s superior performance on accuracy and efficiency,compared with existing distributed learning frameworks.
文摘BACKGROUND： Evidence illustrates that androgen has a neuroprotective role. However, whether androgen also has the protective effect on hippocampal neurons during free radical mediated injury remains unclear. OBJECTIVE： To investigate the neuroprotective effect of androgen on hippocampal neurons during free radical damage. DESIGN, TIME AND SETTING： A controlled in vitro experiment was performed at the Department of Human Anatomy, Cell Culture Lab, and Neuroendocrinology Lab, Basic Medical School, Hebei Medical University from February to June 2009. MATERIALS： Testosterone was provided by Tianjin Jinyao Amino Acid Company, China. METHODS： Primary cultured neurons from 24 Sprague Dawley rats were randomly assigned into four groups： control, H202, testosterone, and testosterone （pre-added） plus H2O2 groups. MAIN OUTCOME MEASURES： The positive cell ratio of microtubule associated protein-Ⅱ and neuron specific enolase was determined by immunocytochemistry. Neuronal morphology was observed by hematoxylin-eosin staining and Nissl staining. Cell vitality and viability were determined using an inverted phase contrast microscope. The content of nitric oxide synthase, malondialdehyde, and superoxide dismutase were measured with a spectrophotometer. RESULTS： As compared with the control group, cell vitality and viability, and superoxide dismutase level were significantly decreased in the H202 group （P 〈 0.05）, while nitric oxide synthase and malondialdehyde levels were significantly increased （P 〈 0.05）. Neuronal vitality and viability as well as superoxide dismutase level in the testosterone plus H2O2 group were significantly greater than in the H2O2 group （P 〈 0.05）, and nitric oxide synthase and malondialdehyde levels were significantly less than in the H2O2 group （P〈 0.05）. CONCLUSION： Androgen partially reversed H2O2-induced neuronal damage and protected neurons.
基金supported by the National Natural Science Foundation of China(Nos.11674269,61975168).
文摘Here we developed a novel wavelength-switchable visible continuous-wave(CW)Pr^(3+):YLF laser around 670 nm.In single-wavelength laser operations,the maximum output powers of 2.60 W,1.26 W,and 0.21 W,the maximum slope efficiencies of 34.7%,27.3%,and 12.3%were achieved with good beam qualities(M^(2)<1.6)at 670.4 nm,674.2 nm,and 678.9 nm,respectively.Record-high output power(2.6 W)and record-high slope efficiency(34.7%)were achieved for the Pr^(3+):YLF laser operation at 670.4 nm.This is also the first demonstration of longer-wavelength peaks beyond 670 nm in the^(3)P_(1)→^(3)F_(3)transition of Pr^(3+):YLF.In multi-wavelength laser operations,the dual-wavelength lasings,including 670.1/674.8 nm,670.1/679.1 nm,and 675.0/679.4 nm,were obtained by fine adjustment of one/two etalons within the cavity.Furthermore,the triple-wavelength lasings,e.g.672.2/674.2/678.6 nm and 670.4/674.8/679.4 nm,were successfully demonstrated.Moreover,both the first-order vortex lasers(LG_(0)^(+1)and LG_(0)^(-1)modes)at 670.4 nm were obtained by off-axis pumping.
基金This research was financially supported by the National Natural Science Foundation of China(Grant No.61379145)the Joint Funds of CETC(Grant No.20166141B020101).
文摘The extreme imbalanced data problem is the core issue in anomaly detection.The amount of abnormal data is so small that we cannot get adequate information to analyze it.The mainstream methods focus on taking fully advantages of the normal data,of which the discrimination method is that the data not belonging to normal data distribution is the anomaly.From the view of data science,we concentrate on the abnormal data and generate artificial abnormal samples by machine learning method.In this kind of technologies,Synthetic Minority Over-sampling Technique and its improved algorithms are representative milestones,which generate synthetic examples randomly in selected line segments.In our work,we break the limitation of line segment and propose an Imbalanced Triangle Synthetic Data method.In theory,our method covers a wider range.In experiment with real world data,our method performs better than the SMOTE and its meliorations.
基金The authors would like to acknowledge the financial support from the National Natural Science Foundation of China(No.61379145)the Joint Funds of CETC(Grant No.20166141B08020101).
文摘By using efficient and timely medical diagnostic decision making,clinicians can positively impact the quality and cost of medical care.However,the high similarity of clinical manifestations between diseases and the limitation of clinicians’knowledge both bring much difficulty to decision making in diagnosis.Therefore,building a decision support system that can assist medical staff in diagnosing and treating diseases has lately received growing attentions in the medical domain.In this paper,we employ a multi-label classification framework to classify the Chinese electronic medical records to establish corresponding relation between the medical records and disease categories,and compare this method with the traditional medical expert system to verify the performance.To select the best subset of patient features,we propose a feature selection method based on the composition and distribution of symptoms in electronic medical records and compare it with the traditional feature selection methods such as chi-square test.We evaluate the feature selection methods and diagnostic models from two aspects,false negative rate(FNR)and accuracy.Extensive experiments have conducted on a real-world Chinese electronic medical record database.The evaluation results demonstrate that our proposed feature selection method can improve the accuracy and reduce the FNR compare to the traditional feature selection methods,and the multi-label classification framework have better accuracy and lower FNR than the traditional expert system.
基金This work is supported by The National Key Research and Development Program of China(2018YFB1800202,2016YFB1000302,SQ2019ZD090149,2018YFB0204301).
文摘With the continuous development of e-commerce,consumers show increasing interest in posting comments on consumption experience and quality of commodities.Meanwhile,people make purchasing decisions relying on other comments much more than ever before.So the reliability of commodity comments has a significant impact on ensuring consumers’equity and building a fair internet-trade-environment.However,some unscrupulous online-sellers write fake praiseful reviews for themselves and malicious comments for their business counterparts to maximize their profits.Those improper ways of self-profiting have severely ruined the entire online shopping industry.Aiming to detect and prevent these deceptive comments effectively,we construct a model of Multi-Filters Convolutional Neural Network(MFCNN)for opinion spam detection.MFCNN is designed with a fixed-length sequence input and an improved activation function to avoid the gradient vanishing problem in spam opinion detection.Moreover,convolution filters with different widths are used in MFCNN to represent the sentences and documents.Our experimental results show that MFCNN outperforms current state-of-the-art methods on standard spam detection benchmarks.
基金National Natural Science Foundation of China(Nos.61475129,11674269)Fundamental Research Funds for the Central Universities(No.20720180057)Natural Science Foundation of Fujian Province for Distinguished Young Scientists(No.2017J06016)。
文摘Mid-infrared(MIR)fiber pulsed lasers are of tremendous application interest in eye-safe LIDAR,spectroscopy,chemical detection and medicine.So far,these MIR lasers largely required bulk optical elements,complex free-space light alignment and large footprint,precluding compact all-fiber structure.Here,we proposed and demonstrated an all-fiberized structured gain-switched Ho3+-doped ZBLAN fiber laser operating around 2.9μm.A home-made 1146 nm Raman fiber pulsed laser was utilized to pump highly concentrated single-cladding Ho3+-doped ZBLAN fiber with different lengths of 2 m or 0.25 m.A home-made MIR fiber mirror and a perpendicular-polished ZBLAN fiber end construct the all-fiberized MIR cavity.Stable gain-switched multiple states with a sub-pulse number tuned from 1 to 8 were observed.The effects of gain fiber length,pump power,pump repetition rate and output coupling ratio on performance of gain-switched pulses were further investigated in detail.The shortest pulse duration of 283 ns was attained with 10 kHz repetition rate.The pulsed laser,centered at 2.92μm,had a maximum average output power of 54.2 mW and a slope efficiency of 10.12%.It is,to the best of our knowledge,the first time to demonstrate a mid-infrared gain-switched Ho3+:ZBLAN fiber laser with compact all-fiber structure.
基金The work is supported by the National Key Research and Development Program of China(2018YFB1800202)the NUDT Research Grants(No.ZK19-38).
文摘Malicious attacks can be launched by misusing the network address translation technique as a camouflage.To mitigate such threats,network address translation identification is investigated to identify network address translation devices and detect abnormal behaviors.However,existingmethods in this field are mainly developed for relatively small-scale networks and work in an offline manner,which cannot adapt to the real-time inference requirements in high-speed network scenarios.In this paper,we propose a flexible and efficient network address translation identification scheme based on actively measuring the distance of a round trip to a target with decremental time-tolive values.The basic intuition is that the incoming and outgoing traffic froma network address translation device usually experiences the different number of hops,which can be discovered by probing with dedicated time-to-live values.We explore a joint effort of parallel transmission,stateless probes,and flexible measuring reuse to accommodate the efficiency of the measuring process.We further accelerate statistical countingwith a new sublinear space data structure Bi-sketch.We implement a prototype and conduct real-world deployments with 1000 volunteers in 31 Chinese provinces,which is believed to bring insight for ground truth collection in this field.Experiments onmulti-sources datasets show that our proposal can achieve as high precision and recall as 95%with a traffic handling throughput of over 106 pps.
基金This work is supported by The National Key Research and Development Program of China(2018YFB1800202,2018YFB0204301,2016YFB1000302,SQ2019ZD090149).
文摘Alias resolution,mapping IP addresses to routers,is a critical step in obtaining a network topology.The latest work on alias resolution is based on special fields in the packet,such as IP ID,port number,etc.However,for security reasons,most network devices block packets for setting options,and some related fields exist only in IPv4,so these methods cannot be used for alias resolution of IPv6.In order to solve the above problems,we propose an alias analysis method based on delay sequence analysis.In this article,we present a new model to describe the distribution of Internet delays and give a mathematical proof.After experimental measurements using the Macroscopic Internet Topology Data Kit(ITDK)and Ark IPv6 Topology Dataset,it was found that the statistical differences in most alias delay models were very small.The statistical differences in the non-alias delay models are spread over a wide range.Using the wavelet decomposition in delay sequence,it was found that the approximate components and the detail components of the delay sequence of aliases were the same after filtering out the noise,which provided a theoretical explanation for the experimental results.This technology is applicable to both IPv4 and IPv6.
文摘For many Internet companies,a huge amount of KPIs(e.g.,server CPU usage,network usage,business monitoring data)will be generated every day.How to closely monitor various KPIs,and then quickly and accurately detect anomalies in such huge data for troubleshooting and recovering business is a great challenge,especially for unlabeled data.The generated KPIs can be detected by supervised learning with labeled data,but the current problem is that most KPIs are unlabeled.That is a time-consuming and laborious work to label anomaly for company engineers.Build an unsupervised model to detect unlabeled data is an urgent need at present.In this paper,unsupervised learning DBSCAN combined with feature extraction of data has been used,and for some KPIs,its best F-Score can reach about 0.9,which is quite good for solving the current problem.
文摘The research and analysis of Internet topology is hot in the field of networkmeasurement, which have important applications in network security, traffic schedulingand many other fields. Most of the existing works are focused on the AS-level and routerleveltopology, but few works are about the IP-level topology. In fact, obtaining thetopology of each continent and knowing how the topologies of the continents areconnected to each other can help us understanding the Internet around the world morethoroughly. In this paper, we obtained data sets from RIPE, constructed and analyzednetwork topologies of all the continents. By analyzing the topological connectionsbetween continents, we found out that most of the junctions of inter-continent traces arelocated in a few countries.
基金This work was financially supported by the National Natural Science Foundation of China(Grant No.61379145)the Joint Funds of CETC(Grant No.20166141B08020101).
文摘While smart devices based on ARM processor bring us a lot of convenience,they also become an attractive target of cyber-attacks.The threat is exaggerated as commodity OSes usually have a large code base and suffer from various software vulnerabilities.Nowadays,adversaries prefer to steal sensitive data by leaking the content of display output by a security-sensitive application.A promising solution is to exploit the hardware visualization extensions provided by modern ARM processors to construct a secure display path between the applications and the display device.In this work,we present a scheme named SecDisplay for trusted display service,it protects sensitive data displayed from being stolen or tampered surreptitiously by a compromised OS.The TCB of SecDisplay mainly consists of a tiny hypervisor and a super light-weight rendering painter,and has only^1400 lines of code.We implemented a prototype of SecDisplay and evaluated its performance overhead.The results show that SecDisplay only incurs an average drop of 3.4%.
基金the China Postdoctoral Science Foundation(2023TQ0089)the National Natural Science Foundation of China(Nos.62072465,62172155)the Science and Technology Innovation Program of Hunan Province(Nos.2022RC3061,2023RC3027).
文摘The global Internet is a complex network of interconnected autonomous systems(ASes).Understanding Internet inter-domain path information is crucial for understanding,managing,and improving the Internet.The path information can also help protect user privacy and security.However,due to the complicated and heterogeneous structure of the Internet,path information is not publicly available.Obtaining path information is challenging due to the limited measurement probes and collectors.Therefore,inferring Internet inter-domain paths from the limited data is a supplementary approach to measure Internet inter-domain paths.The purpose of this survey is to provide an overview of techniques that have been conducted to infer Internet inter-domain paths from 2005 to 2023 and present the main lessons from these studies.To this end,we summarize the inter-domain path inference techniques based on the granularity of the paths,for each method,we describe the data sources,the key ideas,the advantages,and the limitations.To help readers understand the path inference techniques,we also summarize the background techniques for path inference,such as techniques to measure the Internet,infer AS relationships,resolve aliases,and map IP addresses to ASes.A case study of the existing techniques is also presented to show the real-world applications of inter-domain path inference.Additionally,we discuss the challenges and opportunities in inferring Internet inter-domain paths,the drawbacks of the state-of-the-art techniques,and the future directions.
基金This work is supported by the National Key Research and Development Program of China(2022YFF1203001)National Natural Science Foundation of China(Nos.62072465,62102425)the Science and Technology Innovation Program of Hunan Province(Nos.2022RC3061,2023RC3027).
文摘Time series segmentation has attracted more interests in recent years,which aims to segment time series into different segments,each reflects a state of the monitored objects.Although there have been many surveys on time series segmentation,most of them focus more on change point detection(CPD)methods and overlook the advances in boundary detection(BD)and state detection(SD)methods.In this paper,we categorize time series segmentation methods into CPD,BD,and SD methods,with a specific focus on recent advances in BD and SD methods.Within the scope of BD and SD,we subdivide the methods based on their underlying models/techniques and focus on the milestones that have shaped the development trajectory of each category.As a conclusion,we found that:(1)Existing methods failed to provide sufficient support for online working,with only a few methods supporting online deployment;(2)Most existing methods require the specification of parameters,which hinders their ability to work adaptively;(3)Existing SD methods do not attach importance to accurate detection of boundary points in evaluation,which may lead to limitations in boundary point detection.We highlight the ability to working online and adaptively as important attributes of segmentation methods,the boundary detection accuracy as a neglected metrics for SD methods.
基金supported by the National Key Research and Development Program of China(2020YFC2003404)the National Natura Science Foundation of China(No.62072465,62172155,62102425,62102429)+1 种基金the Science and Technology Innovation Program of Hunan Province(Nos.2022RC3061,2021RC2071)the Natural Science Foundation of Hunan Province(No.2022JJ40564).
文摘As an emerging discipline,machine learning has been widely used in artificial intelligence,education,meteorology and other fields.In the training of machine learning models,trainers need to use a large amount of practical data,which inevitably involves user privacy.Besides,by polluting the training data,a malicious adversary can poison the model,thus compromising model security.The data provider hopes that the model trainer can prove to them the confidentiality of the model.Trainer will be required to withdraw data when the trust collapses.In the meantime,trainers hope to forget the injected data to regain security when finding crafted poisoned data after the model training.Therefore,we focus on forgetting systems,the process of which we call machine unlearning,capable of forgetting specific data entirely and efficiently.In this paper,we present the first comprehensive survey of this realm.We summarize and categorize existing machine unlearning methods based on their characteristics and analyze the relation between machine unlearning and relevant fields(e.g.,inference attacks and data poisoning attacks).Finally,we briefly conclude the existing research directions.
基金supported by The National Key Research and Development Program of China(2018YFB1800202,2016YFB1000302,SQ2019ZD090149,2018YFB0204301)the CETC Joint Advanced Research Foundation(6141B08080101)+1 种基金The Major Special Science and Technology Project of Hainan Province(ZDKJ2019008)The New Generation of Artificial Intelligence Special Action Project(AI20191125008).
文摘Deep learning technology has been widely used in computer vision,speech recognition,natural language processing,and other related fields.The deep learning algorithm has high precision and high reliability.However,the lack of resources in the edge terminal equipment makes it difficult to run deep learning algorithms that require more memory and computing power.In this paper,we propose MoTransFrame,a general model processing framework for deep learning models.Instead of designing a model compression algorithm with a high compression ratio,MoTransFrame can transplant popular convolutional neural networks models to resources-starved edge devices promptly and accurately.By the integration method,Deep learning models can be converted into portable projects for Arduino,a typical edge device with limited resources.Our experiments show that MoTransFrame has good adaptability in edge devices with limited memories.It is more flexible than other model transplantation methods.It can keep a small loss of model accuracy when the number of parameters is compressed by tens of times.At the same time,the computational resources needed in the reasoning process are less than what the edge node could handle.
基金supported by the National Natural Science Foundation of China(Nos.62072465,62102425)the Science and Technology Innovation Program of Hunan Province(Nos.2022RC3061,2023RC3027).
文摘In the era of internet proliferation,safeguarding digital media copyright and integrity,especially for images,is imperative.Digital watermarking stands out as a pivotal solution for image security.With the advent of deep learning,watermarking has seen significant advancements.Our review focuses on the innovative deep watermarking approaches that employ neural networks to identify robust embedding spaces,resilient to various attacks.These methods,characterized by a streamlined encoder-decoder architecture,have shown enhanced performance through the incorporation of novel training modules.This article offers an in-depth analysis of deep watermarking’s core technologies,current status,and prospective trajectories,evaluating recent scholarly contributions across diverse frameworks.It concludes with an overview of the technical hurdles and prospects,providing essential insights for ongoing and future research endeavors in digital image watermarking.
文摘Wearable devices are becoming more popular in our daily life.They are usually used to monitor health status,track fitness data,or even do medical tests,etc.Since the wearable devices can obtain a lot of personal data,their security issues are very important.Motivated by the consideration that the current pairing mechanisms of Bluetooth Low Energy(BLE)are commonly impractical or insecure for many BLE based wearable devices nowadays,we design and implement a security framework in order to protect the communication between these devices.The security framework is a supplement to the Bluetooth pairing mechanisms and is compatible with all BLE based wearable devices.The framework is a module between the application layer and the GATT(Generic Attribute Profile)layer in the BLE architecture stack.When the framework starts,a client and a server can automatically and securely establish shared fresh keys following a designed protocol;the services of encrypting and decrypting messages are provided to the applications conveniently by two functions;application data are securely transmitted following another protocol using the generated keys.Prudential principles are followed by the design of the framework for security purposes.It can protect BLE based wearable devices from replay attacks,Man-in-The-Middle attacks,data tampering,and passive eavesdropping.We conduct experiments to show that the framework can be conveniently deployed with practical operational cost of power consumption.The protocols in this framework have been formally verified that the designed security goals are satisfied.
文摘In geometry processing,symmetry research benefits from global geo-metric features of complete shapes,but the shape of an object captured in real-world applications is often incomplete due to the limited sensor resolution,single viewpoint,and occlusion.Different from the existing works predicting symmetry from the complete shape,we propose a learning approach for symmetry predic-tion based on a single RGB-D image.Instead of directly predicting the symmetry from incomplete shapes,our method consists of two modules,i.e.,the multi-mod-al feature fusion module and the detection-by-reconstruction module.Firstly,we build a channel-transformer network(CTN)to extract cross-fusion features from the RGB-D as the multi-modal feature fusion module,which helps us aggregate features from the color and the depth separately.Then,our self-reconstruction net-work based on a 3D variational auto-encoder(3D-VAE)takes the global geo-metric features as input,followed by a prediction symmetry network to detect the symmetry.Our experiments are conducted on three public datasets:ShapeNet,YCB,and ScanNet,we demonstrate that our method can produce reliable and accurate results.
