Search results for: and Wireless sensor networks.

Authors: Qingqing Wang, Shouming Zhong

Abstract:

This paper is concerned with the stability problem with two additive time-varying delay components. By choosing one augmented Lyapunov-Krasovskii functional, using some new zero equalities, and combining linear matrix inequalities (LMI) techniques, two new sufficient criteria ensuring the global stability asymptotic stability of DNNs is obtained. These stability criteria are present in terms of linear matrix inequalities and can be easily checked. Finally, some examples are showed to demonstrate the effectiveness and less conservatism of the proposed method.

Keywords: Neural networks, Globally asymptotic stability, LMI approach, Additive time-varying delays.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1565

Authors: Mario Gonzalez, David Dominguez, Francisco B. Rodriguez

Abstract:

The model of neural networks on the small-world topology, with metric (local and random connectivity) is investigated. The synaptic weights are random, driving the network towards a chaotic state for the neural activity. An ordered macroscopic neuron state is induced by a bias in the network connections. When the connections are mainly local, the network emulates a block-like structure. It is found that the topology and the bias compete to influence the network to evolve into a global or a block activity ordering, according to the initial conditions.

Keywords: Block attractor, random interaction, small world, spin glass.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1336

Authors: S. M. Kamruzzaman, Md. Monirul Islam

Abstract:

Although backpropagation ANNs generally predict better than decision trees do for pattern classification problems, they are often regarded as black boxes, i.e., their predictions cannot be explained as those of decision trees. In many applications, it is desirable to extract knowledge from trained ANNs for the users to gain a better understanding of how the networks solve the problems. A new rule extraction algorithm, called rule extraction from artificial neural networks (REANN) is proposed and implemented to extract symbolic rules from ANNs. A standard three-layer feedforward ANN is the basis of the algorithm. A four-phase training algorithm is proposed for backpropagation learning. Explicitness of the extracted rules is supported by comparing them to the symbolic rules generated by other methods. Extracted rules are comparable with other methods in terms of number of rules, average number of conditions for a rule, and predictive accuracy. Extensive experimental studies on several benchmarks classification problems, such as breast cancer, iris, diabetes, and season classification problems, demonstrate the effectiveness of the proposed approach with good generalization ability.

Keywords: Backpropagation, clustering algorithm, constructivealgorithm, continuous activation function, pruning algorithm, ruleextraction algorithm, symbolic rules.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1615

Authors: Wei Zhang

Abstract:

With the rapid development of deep learning, neural network and deep learning algorithms play a significant role in various practical applications. Due to the high accuracy and good performance, Convolutional Neural Networks (CNNs) especially have become a research hot spot in the past few years. However, the size of the networks becomes increasingly large scale due to the demands of the practical applications, which poses a significant challenge to construct a high-performance implementation of deep learning neural networks. Meanwhile, many of these application scenarios also have strict requirements on the performance and low-power consumption of hardware devices. Therefore, it is particularly critical to choose a moderate computing platform for hardware acceleration of CNNs. This article aimed to survey the recent advance in Field Programmable Gate Array (FPGA)-based acceleration of CNNs. Various designs and implementations of the accelerator based on FPGA under different devices and network models are overviewed, and the versions of Graphic Processing Units (GPUs), Application Specific Integrated Circuits (ASICs) and Digital Signal Processors (DSPs) are compared to present our own critical analysis and comments. Finally, we give a discussion on different perspectives of these acceleration and optimization methods on FPGA platforms to further explore the opportunities and challenges for future research. More helpfully, we give a prospect for future development of the FPGA-based accelerator.

Keywords: Deep learning, field programmable gate array, FPGA, hardware acceleration, convolutional neural networks, CNN.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 892

Authors: M. R. Mustafa, M. H. Isa, R. B. Rezaur

Abstract:

The use of artificial neural network (ANN) modeling for prediction and forecasting variables in water resources engineering are being increasing rapidly. Infrastructural applications of ANN in terms of selection of inputs, architecture of networks, training algorithms, and selection of training parameters in different types of neural networks used in water resources engineering have been reported. ANN modeling conducted for water resources engineering variables (river sediment and discharge) published in high impact journals since 2002 to 2011 have been examined and presented in this review. ANN is a vigorous technique to develop immense relationship between the input and output variables, and able to extract complex behavior between the water resources variables such as river sediment and discharge. It can produce robust prediction results for many of the water resources engineering problems by appropriate learning from a set of examples. It is important to have a good understanding of the input and output variables from a statistical analysis of the data before network modeling, which can facilitate to design an efficient network. An appropriate training based ANN model is able to adopt the physical understanding between the variables and may generate more effective results than conventional prediction techniques.

Keywords: ANN, discharge, modeling, prediction, sediment,

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 5682

Authors: Yi Yu, Vu Xuan Nguyen, Gaoxi Xiao

Abstract:

Opinion formation in complex social networks may exhibit complex system dynamics even when based on some simplest system evolution models. An interesting and important issue is the effects of the initial state on the final steady-state opinion distribution. By carrying out extensive simulations and providing necessary discussions, we show that, while different initial opinion distributions certainly make differences to opinion evolution in social systems without noises, in systems with noises, given enough time, different initial states basically do not contribute to making any significant differences in the final steady state. Instead, it is the basal distribution of the preferred opinions that contributes to deciding the final state of the systems. We briefly explain the reasons leading to the observed conclusions. Such an observation contradicts with a long-term belief on the roles of system initial state in opinion formation, demonstrating the dominating role that opinion mutation can play in opinion formation given enough time. The observation may help to better understand certain observations of opinion evolution dynamics in real-life social networks.

Keywords: Opinion formation, Deffuant model, opinion mutation, consensus making.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 672

Authors: P.S.Prakash, S.Selvan

Abstract:

Over the past several years, there has been a considerable amount of research within the field of Quality of Service (QoS) support for distributed multimedia systems. One of the key issues in providing end-to-end QoS guarantees in packet networks is determining a feasible path that satisfies a number of QoS constraints. The problem of finding a feasible path is NPComplete if number of constraints is more than two and cannot be exactly solved in polynomial time. We proposed Feasible Path Selection Algorithm (FPSA) that addresses issues with pertain to finding a feasible path subject to delay and cost constraints and it offers higher success rate in finding feasible paths.

Keywords: feasible path, multiple constraints, path selection, QoS routing

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1749

Authors: Taiwo. M. Akinmuyisitan, John Cosmas

Abstract:

This paper practically demonstrates the theories and concepts of an Advanced Convolutional Neural Network in the design and development of a scalable artificial intelligence model for the detection of criminal masterminds. The technique uses machine vision algorithms to compute the facial characteristics of suspects and classify actors as criminal or non-criminal faces. The paper proceeds further to compare the results of the error accuracy of two popular custom convolutional pre-trained networks, VGG16 and Resnet50. The result shows that VGG16 is probably more efficient than ResNet50 for the dataset we used.

Keywords: Artificial intelligence, convolutional neural networks, Resnet50, VGG16.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 281

Authors: Mohammed Benbrahim, Khalid Benjelloun, Aomar Ibenbrahim, Adil Daoudi

Abstract:

The automatic classification of non stationary signals is an important practical goal in several domains. An essential classification task is to allocate the incoming signal to a group associated with the kind of physical phenomena producing it. In this paper, we present a modular system composed by three blocs: 1) Representation, 2) Dimensionality reduction and 3) Classification. The originality of our work consists in the use of a new wavelet called "Ben wavelet" in the representation stage. For the dimensionality reduction, we propose a new algorithm based on the random projection and the principal component analysis.

Keywords: Seismic signals, Ben Wavelet, Dimensionality reduction, Artificial neural networks, Classification.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1449

Authors: Malinwo Estone Ayikpa

Abstract:

With the rapid increase of grid-connected PV systems over the last decades, genuine challenges have arisen for engineers and professionals of energy field in the planning and operation of existing distribution networks with the integration of new generation sources. However, the conventional distribution network, in its design was not expected to receive other generation outside the main power supply. The tools generally used to analyze the networks become inefficient and cannot take into account all the constraints related to the operation of grid-connected PV systems. Some of these constraints are voltage control difficulty, reverse power flow, and especially voltage unbalance which could be due to the poor distribution of single-phase PV systems in the network. In order to analyze the impact of the connection of small and large number of PV systems to the distribution networks, this paper presents an efficient optimization tool that minimizes voltage unbalance in three-phase distribution networks with active and reactive power injections from the allocation of single-phase and three-phase PV plants. Reactive power can be generated or absorbed using the available capacity and the adjustable power factor of the inverter. Good reduction of voltage unbalance can be achieved by reactive power control of the PV systems. The presented tool is based on the three-phase current injection method and the PV systems are modeled via an equivalent circuit. The primal-dual interior point method is used to obtain the optimal operating points for the systems.

Keywords: Photovoltaic generation, primal-dual interior point method, three-phase optimal power flow, unbalanced system.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1087

Authors: Hamed.Vezvaei, Sepideh.Ordibeheshti, Mehdi.Ardjmand

Abstract:

Gasoline Octane Number is the standard measure of the anti-knock properties of a motor in platforming processes, that is one of the important unit operations for oil refineries and can be determined with online measurement or use CFR (Cooperative Fuel Research) engines. Online measurements of the Octane number can be done using direct octane number analyzers, that it is too expensive, so we have to find feasible analyzer, like ANFIS estimators. ANFIS is the systems that neural network incorporated in fuzzy systems, using data automatically by learning algorithms of NNs. ANFIS constructs an input-output mapping based both on human knowledge and on generated input-output data pairs. In this research, 31 industrial data sets are used (21 data for training and the rest of the data used for generalization). Results show that, according to this simulation, hybrid method training algorithm in ANFIS has good agreements between industrial data and simulated results.

Keywords: Adaptive Neuro-Fuzzy Inference Systems, GasolineOctane Number, Soft-sensor, Catalytic Naphtha Reforming

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2192

Authors: Aqsa Ali, Aleem Mushtaq, Attaullah Memon, Monna

Abstract:

In this paper, we present a low cost design for a smart glove that can perform sign language recognition to assist the speech impaired people. Specifically, we have designed and developed an Assistive Hand Gesture Interpreter that recognizes hand movements relevant to the American Sign Language (ASL) and translates them into text for display on a Thin-Film-Transistor Liquid Crystal Display (TFT LCD) screen as well as synthetic speech. Linear Bayes Classifiers and Multilayer Neural Networks have been used to classify 11 feature vectors obtained from the sensors on the glove into one of the 27 ASL alphabets and a predefined gesture for space. Three types of features are used; bending using six bend sensors, orientation in three dimensions using accelerometers and contacts at vital points using contact sensors. To gauge the performance of the presented design, the training database was prepared using five volunteers. The accuracy of the current version on the prepared dataset was found to be up to 99.3% for target user. The solution combines electronics, e-textile technology, sensor technology, embedded system and machine learning techniques to build a low cost wearable glove that is scrupulous, elegant and portable.

Keywords: American sign language, assistive hand gesture interpreter, human-machine interface, machine learning, sensing glove.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2728

Authors: M. Owlia, M. H. Azarsa, M. Khabbazan, A. Mirbagheri

Abstract:

Proprioceptive neuromuscular facilitation exercises are a series of stretching techniques that are commonly used in rehabilitation and exercise therapy. Assessment of these exercises for true maneuvering requires extensive experience in this field and could not be down with patients themselves. In this paper, we developed software that uses Microsoft Kinect sensor, a spherical color marker, and real-time image processing methods to evaluate patient’s performance in generating true patterns of movements. The software also provides the patient with a visual feedback by showing his/her avatar in a Virtual Reality environment along with the correct path of moving hand, wrist and marker. Primary results during PNF exercise therapy of a patient in a room environment shows the ability of the system to identify any deviation of maneuvering path and direction of the hand from the one that has been performed by an expert physician.

Keywords: Image processing, Microsoft Kinect, proprioceptive neuromuscular facilitation, upper extremities assessment, virtual reality.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1936

Authors: Hatem Hajri, Mohamed-Cherif Rahal

Abstract:

Both Lidars and Radars are sensors for obstacle detection. While Lidars are very accurate on obstacles positions and less accurate on their velocities, Radars are more precise on obstacles velocities and less precise on their positions. Sensor fusion between Lidar and Radar aims at improving obstacle detection using advantages of the two sensors. The present paper proposes a real-time Lidar/Radar data fusion algorithm for obstacle detection and tracking based on the global nearest neighbour standard filter (GNN). This algorithm is implemented and embedded in an automative vehicle as a component generated by a real-time multisensor software. The benefits of data fusion comparing with the use of a single sensor are illustrated through several tracking scenarios (on a highway and on a bend) and using real-time kinematic sensors mounted on the ego and tracked vehicles as a ground truth.

Keywords: Ground truth, Hungarian algorithm, lidar Radar data fusion, global nearest neighbor filter.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 959

Authors: Y. Ni, K. Sebri

Abstract:

The practical implementation of audio-video coupled speech recognition systems is mainly limited by the hardware complexity to integrate two radically different information capturing devices with good temporal synchronisation. In this paper, we propose a solution based on a smart CMOS image sensor in order to simplify the hardware integration difficulties. By using on-chip image processing, this smart sensor can calculate in real time the X/Y projections of the captured image. This on-chip projection reduces considerably the volume of the output data. This data-volume reduction permits a transmission of the condensed visual information via the same audio channel by using a stereophonic input available on most of the standard computation devices such as PC, PDA and mobile phones. A prototype called VMIKE (Visio-Microphone) has been designed and realised by using standard 0.35um CMOS technology. A preliminary experiment gives encouraged results. Its efficiency will be further investigated in a large variety of applications such as biometrics, speech recognition in noisy environments, and vocal control for military or disabled persons, etc.

Keywords: Audio-Visual Speech recognition, CMOS Smartsensor, On-Chip image processing.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1825

Authors: Dharmendra Chourishi “Maitraya”, Sridevi Seshadri

Abstract:

Bluetooth is a personal wireless communication technology and is being applied in many scenarios. It is an emerging standard for short range, low cost, low power wireless access technology. Current existing MAC (Medium Access Control) scheduling schemes only provide best-effort service for all masterslave connections. It is very challenging to provide QoS (Quality of Service) support for different connections due to the feature of Master Driven TDD (Time Division Duplex). However, there is no solution available to support both delay and bandwidth guarantees required by real time applications. This paper addresses the issue of how to enhance QoS support in a Bluetooth piconet. The Bluetooth specification proposes a Round Robin scheduler as possible solution for scheduling the transmissions in a Bluetooth Piconet. We propose an algorithm which will reduce the bandwidth waste and enhance the efficiency of network. We define token counters to estimate traffic of real-time slaves. To increase bandwidth utilization, a back-off mechanism is then presented for best-effort slaves to decrease the frequency of polling idle slaves. Simulation results demonstrate that our scheme achieves better performance over the Round Robin scheduling.

Keywords: Piconet, Medium Access Control, Polling algorithm, Scheduling, QoS, Time Division Duplex (TDD).

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1699

Authors: Dorota Galeza

Abstract:

This paper argues that networks, such as the ECN and the American network, are affected by certain small events which are inherent to path dependence and preclude the full evolution towards efficiency. It is advocated that the American network is superior to the ECN in many respects due to its greater flexibility and longer history. This stems in particular from the creation of the American network, which was based on a small number of cases. Such a structure encourages further changes and modifications which are not necessarily radical. The ECN, by contrast, was established by legislative action, which explains its rigid structure and resistance to change. This paper is an attempt to transpose the superiority of the American network on to the ECN. It looks at concepts such as judicial cooperation, harmonisation of procedure, peer review and regulatory impact assessments (RIAs), and dispute resolution procedures.

Keywords: Antitrust, Competition, Networks, Path Dependence.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1611

Authors: Philip Symonds, Jon Taylor, Zaid Chalabi, Michael Davies

Abstract:

Average temperatures worldwide are expected to continue to rise. At the same time, major cities in developing countries are becoming increasingly populated and polluted. Governments are tasked with the problem of overheating and air quality in residential buildings. This paper presents the development of a model, which is able to estimate the occupant exposure to extreme temperatures and high air pollution within domestic buildings. Building physics simulations were performed using the EnergyPlus building physics software. An accurate metamodel is then formed by randomly sampling building input parameters and training on the outputs of EnergyPlus simulations. Metamodels are used to vastly reduce the amount of computation time required when performing optimisation and sensitivity analyses. Neural Networks (NNs) have been compared to a Radial Basis Function (RBF) algorithm when forming a metamodel. These techniques were implemented using the PyBrain and scikit-learn python libraries, respectively. NNs are shown to perform around 15% better than RBFs when estimating overheating and air pollution metrics modelled by EnergyPlus.

Keywords: Neural Networks, Radial Basis Functions, Metamodelling, Python machine learning libraries.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2116

Authors: Rachid.Dehini, Abdesselam.Bassou, Brahim.Ferdi

Abstract:

Active Power Filters (APFs) are today the most widely used systems to eliminate harmonics compensate power factor and correct unbalanced problems in industrial power plants. We propose to improve the performances of conventional APFs by using artificial neural networks (ANNs) for harmonics estimation. This new method combines both the strategies for extracting the three-phase reference currents for active power filters and DC link voltage control method. The ANNs learning capabilities to adaptively choose the power system parameters for both to compute the reference currents and to recharge the capacitor value requested by VDC voltage in order to ensure suitable transit of powers to supply the inverter. To investigate the performance of this identification method, the study has been accomplished using simulation with the MATLAB Simulink Power System Toolbox. The simulation study results of the new (SAPF) identification technique compared to other similar methods are found quite satisfactory by assuring good filtering characteristics and high system stability.

Keywords: Artificial Neural Networks (ANN), p-q theory, (SAPF), Harmonics, Total Harmonic Distortion.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2034

Authors: Juha Kettunen

Abstract:

This study analyzes collaborative and networked academic authorship in higher education. The literature review shows evidence that single authorship has made a gradual paradigm shift to joint authorship. The empirical evidence from the Turku University of Applied Sciences indicates that collaborative authorship has notably increased in the last few years. Co-authorship has extended outside the institution to other domestic and international academic organizations. Co-authorship not only increase the merits of academic scholars but builds and maintains networks of research and development. The results of this study help the authors, editors and partners of research and development projects to have a more concrete understanding of how co-authorship has developed and spread beyond higher education institutions.

Keywords: Co-authorship, social networking, higher education, research and development.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1093

Authors: Tarik Rashid, B. Q. Huang, M-T. Kechadi, B. Gleeson

Abstract:

this paper presents an auto-regressive network called the Auto-Regressive Multi-Context Recurrent Neural Network (ARMCRN), which forecasts the daily peak load for two large power plant systems. The auto-regressive network is a combination of both recurrent and non-recurrent networks. Weather component variables are the key elements in forecasting because any change in these variables affects the demand of energy load. So the AR-MCRN is used to learn the relationship between past, previous, and future exogenous and endogenous variables. Experimental results show that using the change in weather components and the change that occurred in past load as inputs to the AR-MCRN, rather than the basic weather parameters and past load itself as inputs to the same network, produce higher accuracy of predicted load. Experimental results also show that using exogenous and endogenous variables as inputs is better than using only the exogenous variables as inputs to the network.

Keywords: Daily peak load forecasting, neural networks, recurrent neural networks, auto regressive multi-context neural network.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2543

Authors: Roza Dastres, Mohsen Soori

Abstract:

In order to electronically exchange information between network users in the web of data, different software such as outlook is presented. So, the traffic of users on a site or even the floors of a building can be decreased as a result of applying a secure and reliable data sharing software. It is essential to provide a fast, secure and reliable network system in the data sharing webs to create an advanced communication systems in the users of network. In the present research work, different encoding methods and algorithms in data sharing systems is studied in order to increase security of data sharing systems by preventing the access of hackers to the transferred data. To increase security in the networks, the possibility of textual conversation between customers of a local network is studied. Application of the encryption and decryption algorithms is studied in order to increase security in networks by preventing hackers from infiltrating. As a result, a reliable and secure communication system between members of a network can be provided by preventing additional traffic in the website environment in order to increase speed, accuracy and security in the network and web systems of data sharing.

Keywords: Secure Socket Layer, Security of networks.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 509

Authors: Ahmad M. Sarhan, Omar I. Al Helalat

Abstract:

In this paper, an Arabic letter recognition system based on Artificial Neural Networks (ANNs) and statistical analysis for feature extraction is presented. The ANN is trained using the Least Mean Squares (LMS) algorithm. In the proposed system, each typed Arabic letter is represented by a matrix of binary numbers that are used as input to a simple feature extraction system whose output, in addition to the input matrix, are fed to an ANN. Simulation results are provided and show that the proposed system always produces a lower Mean Squared Error (MSE) and higher success rates than the current ANN solutions.

Keywords: ANN, Backpropagation, Gaussian, LMS, MSE, Neuron, standard deviation, Widrow-Hoff rule.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2013

Authors: Luke McDonagh

Abstract:

Over the past three decades, free and open source software (FOSS) programmers have developed new, innovative and legally binding licences that have in turn enabled the creation of innumerable pieces of everyday software, including Linux, Mozilla Firefox and Open Office. That FOSS has been highly successful in competing with 'closed source software' (e.g. Microsoft Office) is now undeniable, but in noting this success, it is important to examine in detail why this system of FOSS has been so successful. One key reason is the existence of networks or communities of programmers, who are bound together by a key shared social norm of 'reciprocity'. At the same time, these FOSS networks are not unitary – they are highly diverse and there are large divergences of opinion between members regarding which licences are generally preferable: some members favour the flexible ‘free’ or 'no copyleft' licences, such as BSD and MIT, while other members favour the ‘strong open’ or 'strong copyleft' licences such as GPL. This paper argues that without both the existence of the shared norm of reciprocity and the diversity of licences, it is unlikely that the innovative legal framework provided by FOSS would have succeeded to the extent that it has.

Keywords: Open source, software, licences, reciprocity, networks.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1051

Authors: Alexander Goncharenko, Andrey Denisov, Sergey Alyamkin, Evgeny Terentev

Abstract:

The neural network quantization is highly desired procedure to perform before running neural networks on mobile devices. Quantization without fine-tuning leads to accuracy drop of the model, whereas commonly used training with quantization is done on the full set of the labeled data and therefore is both time- and resource-consuming. Real life applications require simplification and acceleration of quantization procedure that will maintain accuracy of full-precision neural network, especially for modern mobile neural network architectures like Mobilenet-v1, MobileNet-v2 and MNAS. Here we present a method to significantly optimize training with quantization procedure by introducing the trained scale factors for discretization thresholds that are separate for each filter. Using the proposed technique, we quantize the modern mobile architectures of neural networks with the set of train data of only ∼ 10% of the total ImageNet 2012 sample. Such reduction of train dataset size and small number of trainable parameters allow to fine-tune the network for several hours while maintaining the high accuracy of quantized model (accuracy drop was less than 0.5%). Ready-for-use models and code are available in the GitHub repository.

Keywords: Distillation, machine learning, neural networks, quantization.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 730

Authors: K. Shiba, T. Kaburagi, Y. Kurihara

Abstract:

With global aging, people who require care, such as people with dementia (PwD), are increasing within many developed countries. And PwDs may wander and unconsciously set foot outdoors, it may lead serious accidents, such as, traffic accidents. Here, round-the-clock monitoring by caregivers is necessary, which can be a burden for the caregivers. Therefore, an automatic wandering detection system is required when an elderly person wanders outdoors, in which case the detection system transmits a ‘moving’ followed by an ‘absence’ state. In this paper, we focus on the transition from the ‘resting’ to the ‘absence’ state, via the ‘moving’ state as one of the wandering transitions. To capture the transition of the three states, our method based on the hidden Markov model (HMM) is built. Using our method, the restraint where the ‘resting’ state and ‘absence’ state cannot be transmitted to each other is applied. To validate our method, we conducted the experiment with 10 subjects. Our results show that the method can classify three states with 0.92 accuracy.

Keywords: Wander, microwave Doppler sensor, respiratory frequency band, the state transition, hidden Markov model.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 848

Authors: Sun-Hee Han, Hun-Jung Lim, Tai-Myoung Chung

Abstract:

Vehicular Ad-Hoc Networks (VANET) can provide communications between vehicles or infrastructures. It provides the convenience of driving and the secure driving to reduce accidents. In VANET, the security is more important because it is closely related to accidents. Additionally, VANET raises a privacy issue because it can track the location of vehicles and users- identity when a security mechanism is provided. In this paper, we analyze the problem of an existing solution for security requirements required in VANET, and resolve the problem of the existing method when a key management mechanism is provided for the security operation in VANET. Therefore, we show suitability of the Long Term Evolution (LTE) in VANET for the solution of this problem.

Keywords: VANET, Privacy, Security, LTE

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1814

Authors: D. SuganyaDevi, G. Padmavathi

Abstract:

Reliable secure multicast communication in mobile adhoc networks is challenging due to its inherent characteristics of infrastructure-less architecture with lack of central authority, high packet loss rates and limited resources such as bandwidth, time and power. Many emerging commercial and military applications require secure multicast communication in adhoc environments. Hence key management is the fundamental challenge in achieving reliable secure communication using multicast key distribution for mobile adhoc networks. Thus in designing a reliable multicast key distribution scheme, reliability and congestion control over throughput are essential components. This paper proposes and evaluates the performance of an enhanced optimized multicast cluster tree algorithm with destination sequenced distance vector routing protocol to provide reliable multicast key distribution. Simulation results in NS2 accurately predict the performance of proposed scheme in terms of key delivery ratio and packet loss rate under varying network conditions. This proposed scheme achieves reliability, while exhibiting low packet loss rate with high key delivery ratio compared with the existing scheme.

Keywords: Key Distribution, Mobile Adhoc Network, Multicast and Reliability.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1636

Authors: Farah Diyana Abdul Rahman, Wajdi Al-Khateeb, Aisha Hassan Abdalla Hashim

Abstract:

Optical networks are high capacity networks that meet the rapidly growing demand for bandwidth in the terrestrial telecommunications industry. This paper studies and evaluates singlemode and multimode fiber transmission by varying the distance. It focuses on their performance in LAN environment. This is achieved by observing the pulse spreading and attenuation in optical spectrum and eye-diagram that are obtained using OptSim simulator. The behaviors of two modes with different distance of data transmission are studied, evaluated and compared.

Keywords: Attenuation, eye diagram, fiber transmissions, multimode fiber, pulse dispersion, OSNR, single-mode fiber.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2519

Authors: Ahmet Faruk Akyuz, Hasan Sakir Bilge

Abstract:

Depth estimation using stereo images is a challenging problem in computer vision. Many different studies have been carried out to solve this problem. With advancing machine learning, tackling this problem is often done with neural network-based solutions. The images used in these studies are mostly in the visible spectrum. However, the need to use the Infrared (IR) spectrum for depth estimation has emerged because it gives better results than visible spectra in some conditions. At this point, we recommend using thermal-thermal (IR) image pairs for depth estimation. In this study, we used two well-known networks (PSMNet, FADNet) with minor modifications to demonstrate the viability of this idea.

Keywords: thermal stereo matching, depth estimation, deep neural networks, CNN

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 692