We perform a large-scale study to quantify just how severe the privacy leakage problem is in Facebook. As case study, we focus on estimating birth year, which fundamental human attribute and, for many people, private one. Specifically, attempt estimate year of over 1 million Facebook users New York City. examine accuracy estimation procedures several classes users: (i) highly users, who do not make their friend lists public; (ii) hide years but public. To users' ages, exploit underlying...

The object detection based on deep learning is an important application in technology, which characterized by its strong capability of feature and representation compared with the traditional methods. paper first makes introduction classical methods detection, expounds relation difference between detection. Then it introduces emergence elaborates most typical nowadays via learning. In statement methods, focuses framework design working principle models analyzes model performance real-time...

Although emotion recognition techniques have been well developed, the understanding of neural mechanism remains rudimentary. The traditional static network approach cannot reflect entire brain activity at time scale. Instead, a newly introduced temporal is an optimal which can be used to investigate dynamic functional connectivity (FC) human in different states considering time-varying regional interaction. In this study, we focused on and FC analysis with SEED dataset. First, multiband...

Accurate and reliable traffic flow prediction is critical to the safe stable deployment of intelligent transportation systems. However, it very challenging since complex spatial temporal dependence flows. Most existing works require information network structure human intervention model spatial-temporal association data, resulting in low generality unsatisfactory performance. In this paper, we propose a general graph attention based dynamic convolutional (GAGCN) predict flow. GAGCN uses...

A multi-view object detection approach based on deep learning is proposed in this paper. Classical methods regression models are introduced, and the reasons for their weak ability to detect small objects analyzed. To improve performance of these methods, a proposed, model structure working principles explained. Additionally, retrieval accuracy both corresponding classical evaluated compared test dataset. The experimental results show that terms capability, Multi-view YOLO (You Only Look...

Motor imagery (MI) based brain-computer interface (BCI) is a research hotspot and has attracted lots of attention.Within this topic, multiple MI classification challenge due to the difficulties caused by time-varying spatial features across different individuals.To deal with challenge, we tried fuse brain functional connectivity (BFC) one-versus-the-rest filter-bank common pattern (OVR-FBCSP) improve robustness classification.The BFC were extracted phase locking value (PLV), representing...

Motor imagery (MI) is a typical BCI paradigm and has been widely applied into many aspects (e.g. brain-driven wheelchair motor function rehabilitation training). Although significant achievements have achieved, multiple decoding still unsatisfactory. To deal with this challenging issue, firstly, segment of electroencephalogram was extracted preprocessed. Secondly, we filter bank common spatial pattern (FBCSP) one-vs-rest (OVR) strategy to extract the spatio-temporal-frequency features MI....

The increasing popularity of blockchain-based cryptocurrencies has made scalability a primary and urgent concern. Compared with the completely open, uncontrolled public blockchain system, private can provide better access control management. This paper proposes an architecture for distributed blockchain. At same time, we propose three strategies to improve blockchain: optimization block construction, size time optimization, transaction security mechanism optimization. system also redesigns...

Attack graphs play important roles in analyzing network security vulnerabilities, and previous works have provided meaningful conclusions on the generation measurement of attack graphs. However, it is still hard for us to understand a large network, few suggestions been proposed prevent inside malicious attackers from attacking networks. To address these problems, we propose novel approach generate describe Firstly, construct two-layer graph, where upper layer hosts access graph lower...

The regular patterns of soil erosion tend to change at different scales observation, affecting the mechanism and its evolution characteristics. This phenomenon has essential scientific significance for rational allocation land resources studies on sustainable ecosystems. As an important agricultural area in China, Danjiangkou reservoir is threatened by severe erosion. In this study, we selected four kinds landscape pattern metrics, including patch density, fractal dimension, Shannon...

The guaranteed delivery model is widely used in online advertising. publisher sells impressions advance by promising to serve each advertiser an agreed-upon number of target that satisfy specific attribute requirements over a fixed time period. Previous efforts usually the service as crowd-level or user-level supply allocation problem and focus on searching optimal for serving, assuming forecasts are available contracts already signed. Existing techniques not sufficient meet needs today's...

As computational ghost imaging is widely used in the military, radar, and other fields, its security efficiency became more important. In this paper, we propose a compressive encryption scheme based on hyper-chaotic system, DNA encoding, KSVD algorithm for first time. First, 4-dimensional system to generate four long pseudorandom sequences diffuse with operation get phase mask sequence, then <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M1"><mml:mi>N</mml:mi></mml:math>...

In order to study the characteristics and ecological risks of heavy metal pollution in Dongting Lake, 24 sediment cores were collected August, 2017. The content status metals area analyzed by method Igeo, RI mPEC-Q, interrelation source heaty evalutated CA PCA. results showed that concentrations surface sediments followed Cr > Pb Cu As Cd Hg, contents Cr, Cd, Pb, 7.0, 2.0, 9.3, 2.2, 2.3 times greater than background values, respectively. using geo-accumulation index potential risk assessment...

How to choose a chaotic sequence construct S-box is key factor determine the performance of S-box. In this paper, construction method for based on fireworks algorithm proposed, in which used search optimal solution space Lorenz system. The performances are tested through nonlinearity, differential uniformity, strict avalanche criterion, output bits independence criterion and bijectivity, compared with existing classical S-boxes results indicate that paper has more excellent cryptographic...

In order to make recommendations a user, recommender mainly uses two approaches: content-based-filtering approach and collaborative filtering approach. However, they both still have some shortcomings technically. The content-based is difficult handle feature extraction as well user intension prediction. faces the hard issue of cold start problem matrix sparsity problem. this paper, we present an novel hybrid recommendation based on Ontology Neural Network in movie domain. combines...

Content pollution is pervasive in the current peer-to-peer file sharing systems. Many previous reputation models have been proposed to address this problem, however, such strongly rely on participants' feedback.

Nowadays, RFID technology is increasingly become popular and begin to enter many spheres of everyday life. In this paper, we present LocSafe, a "missed-connections" service with privacy grantees based on technology, in order prove an encounter sharing among users the past. LocSafe comprised three parts: Tags, LE Collectors, Social Service Provider. We use detect encounters, attribute-based encryption broadcast establish trust protect users' privacy. evaluate by study problems analysis system...

This Radio Frequency Identification (RFID) systems suffer from different security and privacy problems, among which relay attack is a hot topic recently. A type of related to man-in-the-middle replay attacks, in an attacker relays verbatim message the sender valid receiver message. The may not be aware even sending attacker. main countermeasure against use distance bounding protocols measuring round-trip time between reader tag. In this paper, we consider modification these using `error...

Private attributes of Online Social Network (OSN) users can be inferred from other information (which is usually users' friends and group information). To address this, social networking sites allow to hide their friend lists lists, so that general public cannot see them. However, if a user doesn't make his list public, but have where we find him, do reverse lookup extend the user. Furthermore, many networks non-group members groups (e.g., Face book). These are strong violations OSN privacy,...

P300 has been used in Brain-computer interface (BCI) as a typical visual event-related potential (ERP). However, electroencephalogram (EEG) analysis with large number of channels will cause information redundancy and increase computational complexity. In this paper, combination random forest (RF) genetic algorithm (GA) for EEG channel selection was proposed. Competition III Data set II evaluation the proposed improved based on six-layer batch normalized convolution neural network (BN-CNN)....

Aiming at the contradiction between accuracy and speed in industrial object detection, this paper proposes an detection method based on improved CenterNet. The uses ResNet-50 as Backbone to boost speed, upsampling layer is added feature processing network improve accuracy. expermient results show that mAP of reaches 87.41 %, which 3.44% higher than CenterNet-Res101 method, 31 FPS, 4 FPS faster method.

in current cloud services, users put their data and resources into the so as to enjoy on-demand high quality applications services.Different from conventional services lose control of which is instead manipulated by large-scale cloud.Therefore, service providers (CSP) guarantee that they provide confidence accuracy integrity.Traditional penetration test carried out manually has low efficiency.In this paper, we propose FPTC, a novel framework environment.In there are managers, executors...