Effective convolutional features play an important role in saliency estimation but how to learn powerful for is still a challenging task. FCN-based methods directly apply multi-level without distinction, which leads sub-optimal results due the distraction from redundant details. In this paper, we propose novel attention guided network selectively integrates contextual information progressive manner. Attentive generated by our can alleviate of background thus achieve better performance. On...

Recently, Variational Autoencoders (VAEs) have been successfully applied to collaborative filtering for implicit feedback. However, the performance of resulting model depends a lot on expressiveness inference and latent representation is often too constrained be expressive enough capture true posterior distribution. In this paper, novel framework named VAEGAN proposed address above issue. VAEGAN, we first introduce Adversarial Bayes (AVB) train with arbitrarily model. By utilizing Generative...

The development of fifth-generation (5G) mobile communication technology has become a major driver to the growth Internet Things (IoT) applications. As promising networking paradigm, software-defined (SDN) makes IoT more flexible and agile by decoupling control plane from data plane. With large number heterogeneous devices accessing network, we need divide network into several domains each domain is managed an SDN controller. Controllers share topologies with other form global view entire...

A cloud computing paradigm enables users to access services, applications, and infrastructure resources by using thin clients anywhere at any time. In this paradigm, multiple can share resources. The application or service requests from a user be abstracted as virtual network (VN) request submitted the cloud-based data centers. How map VN onto is challenging issue in resource provisioning. Thus, efficient mapping techniques that intelligently use of are important necessary. Current research...

Software defined network (SDN) can effectively improve the performance of traffic engineering and will be widely used in backbone networks. Therefore, new energy-saving schemes must take SDN into consideration; this action is extremely important owing to rapidly increasing energy consumption telecom Internet service provider (ISP) Meanwhile, introduction current networks incremental most cases, for technical economic reasons. During period, operators manage hybrid which traditional protocols...

Although a restoration approach derives reroute path when failure occurs and greatly reduces forwarding rules in switches compared with protection approach, software-defined networks (SDNs) induce long recovery process because of frequent flow operations between the SDN controller switches. Accordingly, it is indispensable to design new resilience balance time rule occupation. To this end, we leverage flexible aggregation fast solve problem. In proposed each disrupted traffic reassigned...

Edge computing is an indispensable technology that overcomes delay limitations of cloud computing. In edge computing, computational resources are deployed at the network edge, and tasks data end terminals can be efficiently processed by nodes. Considering resource nodes, collaborative integrates nodes provides more efficient services for terminals. This article considers a computation offloading problem in networks, where allocation optimized means load shedding approach: terminal offload...

Abstract Intrusion detection systems help network administrators prepare for and deal with security attacks. These collect information from a variety of sources, analyze them signs intrusion misuse. A techniques have been employed analysis ranging traditional statistical methods to new data mining approaches. In this study the performance three in detecting is examined. An experimental design (3times2x2) created evaluate impact methods, two representation formats, proportion schemes on...

In this letter, we propose a local fast reroute (LFR) algorithm with flow aggregation in software defined networks (SDN). LFR, if link failure is detected, all traffic flows affected by the are aggregated to be new "big" flow. Then, path dynamically deployed SDN controller for LFR reduces number of operations between and switches. Numerical results show that enables recovery while minimizing total entries SDN.

Abstract The intrinsic similarity between camouflaged objects and background environment impedes the automatic detection/segmentation of objects, novel network architectures for deep learning are promising to overcome this challenge improve detection accuracy. However, these existing distinguishing their backgrounds do not account constraint speed, which results in high computational complexity inability meet requirements rapid detection. Therefore, based on human visual system, study...

The rapid urbanization development has brought huge opportunities for the of architectural landscape design, and naturally there will be various challenges accordingly. What this kind speed presents to people is emergence new urban areas, endless construction projects, so on. This paper conducts research on design based data fusion computer virtual reality technology. purpose use dual guidance technology demonstrate feasibility creating a more planned design. article uses SketchUp software...

In this paper, we propose a visual saliency detection algorithm to explore the fusion of various models in manner bootstrap learning. First, an original bootstrapping model, which combines both weak and strong models, is constructed. image priors are exploited generate provides training samples for model. Then, classifier learned based on extracted from We use classify all salient non-salient superpixels input image. To further improve performance, multi-scale maps model integrated,...

To accommodate heterogeneous and sophisticated network services, Network Function Virtualization (NFV) is invented as a hopeful networking technology. The most distinct feature of NFV that it separates functions from physical hardware. In the architecture, various types Virtual Functions (VNFs) are placed on specific software-based middleboxes by telecom providers. Traffic traverses through sequence in pre-defined order, which named Service Chain (SFC). However, how to effectively place VNFs...

A Ku Band (15~18 GHz) 4-Element (4 Transmitters/4 Receivers) fully differential phased array transceiver is designed and fabricated using a 180nm CMOS process. The proposed integrated with T/R switches SPI controller based on an all-RF structure. TX RX channels are placed side-by-side to improve integration density isolation. Each channel consists of 5-bit phase shifter 4-bit attenuator. measured maximum gain 21 dB for 10.8 channel. minimum noise figure switch 9.9 dB. input referred PidB...

Network Function Virtualization (NFV) is an emerging network architecture, which decouples the software implementation of functions from underlying hardware. Telecom operators widely place diverse types Virtual Functions (VNFs) on specified middlebox. Traffic needs to go through a set ordered VNFs forms Service Chain (SFC). However, how efficiently at various locations while minimizing energy consumption still open problem. To this end, we study joint optimization VNF placement and traffic...

With the increasing applications of unmanned aerial vehicles (UAVs) in surveying, mapping, rescue, etc., security autonomous flight complex environments becomes a crucial issue. Deploying UAVs typically requires them to have accurate dynamic obstacle perception, such as detection birds and other flying at high altitudes, well humans ground low altitudes or indoors. This work’s primary goal is cope with both static moving obstacles environment by developing new framework for UAV planning...

ABSTRACT An attempt at using X-ray computerized tomography (CT) and stereology method to estimate the gradation of asphalt mixtures is made. Using CT which a nondestructive test technique obtain batch slice images, analyzing these two-dimensional (2-D) applying probability statistical theory, three- dimensional (3-D) volumetric coarse aggregate under assumption spherical can be estimated from 2-D area gradation. There error between gradations mechanical sieving simple converting that...

Network Function Virtualization (NFV) is an emerging network resource utilization approach which decouples functions from proprietary hardware. To accommodate Service Chain (SFC) requests, service providers offer Virtual (VNF) instances in operator networks. However, how to efficiently place VNFs at various locations while jointly optimizing computing and communication still open problem. this end, we study the resource-efficient VNF placement problem We firstly formulate as Integer Linear...