The rapidly-rising field of machine learning, including deep has inspired applications across many disciplines. In medical imaging, learning been primarily used for image processing and analysis. this paper, we integrate a convolutional neural network (CNN) into the computed tomography (CT) reconstruction process. Our first task is to monitor quality CT images during iterative decide when stop process according an intelligent numerical observer instead using traditional stopping rule, such...

Image aesthetics assessment is essential to multimedia applications such as image retrieval, and personalized search recommendation. Primarily relying on visual information manually-supplied ratings, previous studies in this area have not adequately utilized higher-level semantic information. We incorporate additional textual phrases from user comments jointly represent utilizing multimodal Deep Boltzmann Machine. Given an image, without requiring any associated comments, the proposed...

Running multiple virtual networks over a shared physical network is promising way to support diverse applications, consequently virtualization viewed as the keystone of next-generation architecture. However, decoupling role traditional ISPs into Infrastructure Providers (InPs) and Service (SPs), also brings some new challenges us. For example, how fairly efficiently share sacred resources InPs among SPs key problem. The interaction between SPs, such cooperation competition, makes this topic...

The difficulty of getting medical treatment is one major livelihood issues in China. Since patients lack prior knowledge about the spatial distribution and capacity hospitals, some hospitals have abnormally high or sporadic population densities. This paper presents a new model for estimating spatiotemporal density each hospital based on location-based service (LBS) big data, which would be beneficial to guiding dispersing outpatients. To improve estimation accuracy, several approaches are...

Today's smart devices like fitness tracker, smartwatch, etc., often employ Bluetooth Low Energy (BLE) for data transmission. Such thus become our information portal, e.g., SMS message and notifications are delivered to those through BLE. In this study, we present BlueDoor, which can obtain unauthorized from via BLE vulnerability. We thoroughly examine the protocol, leverage its intrinsic properties designed low-cost embedded wearable bypass encryption authentication in By mimicking a low...

Recently network virtualization is considered as a powerful way to overcome the current ossification of Internet, and viewed keystone next-generation architecture. In virtualization, virtual (VN) embedding problem, which assigns nodes links physical network, plays an important role. Due scarcity resource, efficient effective VN schemes are necessary. Researchers usually narrow down problem space with certain restrictions propose solutions, or pay more attention improve performance link...

Diffuse optical tomography (DOT) is a relatively low cost and portable imaging modality for reconstruction of properties in highly scattering medium, such as human tissue. The inverse problem DOT ill-posed, making high-quality image critical challenge. Because the nature sparsity DOT, regularization has been utilized to achieve reconstruction. However, conventional approaches using sparse optimization are computationally expensive have no selection criteria optimize parameter. In this paper,...

Nowadays, more and Internet of Things (IoT) devices are exposed to the face increasing threats. These always have low computing resources no antivirus installed. Honeypot is widely used collect predict potential Traditional honeypots designed for a specific device, however, cannot work well large amount even emerging diverse IoT devices. We propose Chameleon, an adaptive honeypot Given target Chameleon forwards received network traffic device while ensuring security. stores requests...

The potential scope of children's participation in China is enormous, with one-fifth the world's population there. Processes economic and social change over past two decades have opened up new opportunities for participation, but increased pressures on time. Welfare educational systems are changing a step behind pace reform its consequent increasing inequalities. Traditional ideas childhood, particular subservience to adults, continued despite shifts environment, mitigate against which has...

The current Internet is a wonderful success today and becomes the critical infrastructure for global communication. However, its existing emerging limitations compel network research community to solve various problems on fundamental architecture view next generation Internet. In this paper, we focus in management, control measurement, propose new virtualization-based architecture, which achieves cognitive management direct by introducing knowledge plane. Moreover, proposed designed...

With the development of wind power generation, turbine tripping has become an effective emergency control measure. Its mechanism a research focus. On basis EEAC theory and influence on system transient stability, this article indicates for rotor-angle stability synchronous generators. At time control, leads to change generators, such as mutation electromagnetic variation retardation area. As consequence these change, angle is suffered. In context, analytic expression derived. The factors...

In this paper, we present our RS-SLAM algorithm for monocular camera where the proposal distribution is derived from 5-point RANSAC and image feature measurement uncertainties instead of using easily violated constant velocity model. We propose to do another sampling within all inliers that have best score check inlier misclassifications in original correspondences use hypotheses generated these consensus sets distribution. This mitigate data association errors (inlier misclassifications)...

<title>Abstract</title> In autonomous driving, traffic sign detection is easily affected by environmental lighting, changes in target size, and complex backgrounds. To address these issues, this paper proposes a multi-scale algorithm, SS-YOLO (Small Sign). This method enhances feature fusion generating additional high-resolution layers optimizes small object removing the large-object head. At same time, C3 module replaced with more efficient C3-DS module, EMA attention mechanism introduced...

A 5GHz monolithic fractional-N frequency synthesizer is implemented in 0.18-/spl mu/m CMOS technology. In this synthesizer, the center of VCO twice that required LO and quadrature generated through a divide-by-2 circuit. 3-b third-order /spl Delta//spl Sigma/ modulator employed to randomize division ratio phase noise performance improved. Simulation result shows whole less than -125dBc/Hz (3MHz) settling time 25/spl mu/s. Having resolution 8Hz, meets specifications WLAN 802.11b/g Bluetooth....

Reinforcement Learning (RL) methods are relatively new in the field of aerospace guidance, navigation, and control. This dissertation aims to exploit RL improve autonomy online learning systems with respect a priori unknown system environment, dynamical uncertainties, partial observability. In first part this dissertation, incremental Approximate Dynamic Programming (iADP) proposed. Instead using nonlinear function approximators approximate true cost-to-go, iADP use an (extended) model deal...

摘要: 软件定义网络是近年来的研究热点，流表更新问题则是其中的一个重要问题。该文提出基于分类的流表更新一致性方案，在保证更新一致性的同时，具有通用性强、有效减轻控制负载等特点。文中引入逻辑证明来验证所提方案能够保证流表更新一致性。多个场景的仿真结果显示，与相关研究相比，该文方案有更好的通用性，更新时间基本一致，能有效降低控制负载。 关键词: 软件定义网络 / 流表更新 流表更新一致性

Many techniques have been proposed to database size estimation. However, the emergency of cloud computing introduces new opportunities along with challenges. In cloud, a monitoring proxy can be set up by service provider due ownership infrastructure. The collected data allows for estimate which may black-box them. We claim that relationship between query performance and captured complexity function. One leverage such function table if given execution time. this paper, we propose fine grained...

In network virtualisation, which enables multiple virtual routers to share one physical router, how ensure the scalability and performance of concurrent is a challenging problem in router design. Since only has limited memory resource, it significant efficiently store forwarding tables for routers. Motivated by idea diminishing dissimilarities tables, authors propose novel trie shifting scheme with depth adjusting, finally obtain memory-efficient shared all tables. this scheme, dissimilarity...

In this paper, a new method for edge detection in multispectral images is presented. The based on the assumption that edges correspond to boundary separates two adjacent homogeneous regions, and characterization of are conducted by evaluating difference between neighborhood regions derive intensity orientation. To demonstrate efficiency proposed approach, we carried out computer experiments multiple-band comparison with other methods. Experiments variety have shown it consistent reliable...

In this paper, a 4.8 GHz fully integrated low power, phase noise phase-locked loop (PLL) frequency synthesizer ready for WLAN application is presented. It implemented in 0.18 mum IP6M CMOS process. The chip consumes only 16 mA (including local buffers to the receiver and transmitter) from 1.8 V supply occupies an area of 1.85times1.1 mm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> . From measurements, its in-band -55.8 dBc/Hz...