abstract
stringlengths
5
10.1k
authors
stringlengths
9
1.96k
title
stringlengths
5
367
__index_level_0__
int64
1
1,000k
Corrigendum: `` The Area-Time Complexity of Binary Multiplication''
['Richard P. Brent', 'H. T. Kung']
Corrigendum: `` The Area-Time Complexity of Binary Multiplication''
184,705
In this paper, a novel adaptive tracking controller is proposed for mobile robots in presence of wheel slip and external disturbance force based on neural networks with online weight updating laws. The uncertainties due to the wheel slip and external force are compensated online by neural networks in order to achieve the desired tracking performance. The online weight updating laws are modified versions of the backpropagation with an e-modification term added for robustness. The global uniformly ultimately bounded stability of the system to an arbitrarily small neighborhood of the origin is proven using Lyapunov method. The validity of the proposed controller is confirmed by two simulation examples of tracking a straight line and a U-shape trajectory.
['Ngoc Bach Hoang', 'Hee-Jun Kang']
Neural network-based adaptive tracking control of mobile robots in the presence of wheel slip and external disturbance force
568,689
In this paper, we present a novel micromachined hot-film flow sensor system realized by a technique using a film depositing processes and incorporating a standard printed circuit. Sensor electrodes and electronic circuits are preprinted on a flexible substrate of polyimide (PI), i.e., a flexible printed circuit board (FPCB). The sensing element, which is made of Cr/Ni/Pt with a temperature coefficient of resistance around 2,000 ppm/K, is fabricated on the FPCB by either magnetron sputtering technology or pulsed laser deposition (PLD). The sensor can be packed efficiently at high-density and integrated with signal processing circuits without additional pads. A simple fabrication process using mature technique and materials selection guarantees that the time and costs are greatly reduced. Both steady-state and transient characteristics of the sensors are experimentally tested, and the results presented to validate the effectiveness of the sensors.
['Peng Liu', 'Rong Zhu', 'Ruiyi Que']
A Flexible Flow Sensor System and Its Characteristics for Fluid Mechanics Measurements
541,609
Network lifetime and throughput are one of the prime concerns while designing routing protocols for wireless sensor networks (WSNs). However, most of the existing schemes are either geared towards prolonging network lifetime or improving throughput. This paper presents an energy efficient routing scheme for throughput improvement in WSN. The proposed scheme exploits multilayer cluster design for energy efficient forwarding node selection, cluster heads rotation and both inter- and intra-cluster routing. To improve throughput, we rotate the role of cluster head among various nodes based on two threshold levels which reduces the number of dropped packets. We conducted simulations in the NS2 simulator to validate the performance of the proposed scheme. Simulation results demonstrate the performance efficiency of the proposed scheme in terms of various metrics compared to similar approaches published in the literature.
['Sohail Jabbar', 'Abid Ali Minhas', 'Muhammad Imran', 'Shehzad Khalid', 'Kashif Saleem']
Energy efficient strategy for throughput improvement in wireless sensor networks.
4,516
Most of existing learning-based methods for query-by-example take the query examples as ldquopositiverdquo and build a model for each query. These methods, referred to as query-dependent, only achieved limited success as they can hardly be applied to real-world applications, in which an arbitrary query is usually given. To address this problem, we propose to learn a query-independent model by exploiting the relevance information which exists in the pair of query-document. The proposed approach takes a query-document pair as a sample and extracts a set of query-independent textual and visual features from each pair. It is general and suitable for a real-world video search system since the learned relevance relation is independent on any query. We conducted extensive experiments over TRECVID 2005-2007 corpus and shown superior performance (+37% in Mean Average Precision) to the query-dependent learning approaches.
['Yuan Liu', 'Tao Mei', 'Guo-Jun Qi', 'Xiuqing Wu', 'Xian-Sheng Hua']
Query-independent learning for video search
362,834
We consider a discrete optimization via simulation problem with stochastic constraints on secondary performance measures where both objective and secondary performance measures need to be estimated by simulation. To solve the problem, we present a method called penalty function with memory (PFM), which determines a penalty value for a solution based on history of feasibility check on the solution. PFM converts a DOvS problem with stochastic constraints into a series of new optimization problems without stochastic constraints so that an existing DOvS algorithm can be applied to solve the new problem.
['Chuljin Park', 'Seong-Hee Kim']
Handling stochastic constraints in discrete optimization via simulation
23,778
More than a decade has passed since Herbrand’s theorem was first generalized to arbitrary institutions, enabling in this way the development of the logic-programming paradigm over formalisms beyond the conventional framework of relational first-order logic. Despite the mild assumptions of the original theory, recent developments have shown that the institution-based approach cannot capture constructions that arise when service-oriented computing is presented as a form of logic programming, thus prompting the need for a new perspective on Herbrand’s theorem founded instead upon a concept of generalized substitution system. In this paper, we formalize the connection between the institution- and the substitution-system-based approach to logic programming by investigating a number of features of institutions, like the existence of a quantification space or of representable substitutions, under which they give rise to suitable generalized substitution systems. Building on these results, we further show how the original institution independent versions of Herbrand’s theorem can be obtained as concrete instances of a more general result.
['Ionut Tutu', 'José Luiz Fiadeiro']
Revisiting the Institutional Approach to Herbrand’s Theorem
695,420
How Many Beans Make Five? The Consensus Problem in Music-Genre Classification and a New Evaluation Method for Single-Genre Categorisation Systems.
['Alastair J. D. Craft', 'Geraint A. Wiggins', 'Tim Crawford']
How Many Beans Make Five? The Consensus Problem in Music-Genre Classification and a New Evaluation Method for Single-Genre Categorisation Systems.
770,680
Train control systems like most digital controllers are, by definition, hybrid systems as they interact with or try to control some aspects of the physical world. Detailed behavior modeling with constraints specifica-tion and formal verification, required for reliability prediction, is a great challenge for hybrid system designers. Train control systems further intensify this challenge with extensive interaction between computing units and their physical environment and their mutual dependence on each other. In this paper, we investigate behavior modeling and formal verification of Chinese Train Control System Level 3 (CTCS-3) using Architectural Anal-ysis & Design Language (AADL) to cope with this challenge. AADL is an architecture description language for embedded systems and is based on model-based engineering paradigm. Along with structural modeling of em-bedded systems using the core language constructs, AADL also provides support for language extension through annex sublanguages. In system requirements specification document, the behavior of the CTCS-3 is specified as a set of basic operation scenarios that cooperate with each other to achieve safe and secure functionality of trains. Movement Authority (MA) scenario, explored in this paper, is considered as a basic and most crucial scenario to prevent trains from colliding with each other. The detailed discrete behavior of control system is modeled and verified using the Behavior Language for Embedded Systems with Software (BLESS) annex sublanguage of AADL, and the continuous behavior of train with the cyber-physical interaction (communication between train and control system) is modeled using the Hybrid annex sublanguage. The behavior of the MA scenario at system level is verified using the Hybrid Hoare Logic theorem prover. Behavior constraints are specified as assertions using first-order logic formulas augmented with a simple temporal operator.
['Ehsan Ahmad', 'Yunwei Dong', 'Brian R. Larson', 'Jidong Lü', 'Tao Tang', 'Naijun Zhan']
Behavior modeling and verification of movement authority scenario of Chinese Train Control System using AADL
561,754
A new large margin classifier, named Maxi-Min Margin Machine (M 4 ) is proposed in this paper. This new classifier is constructed based on both a "local: and a "global" view of data, while the most popular large margin classifier, Support Vector Machine (SVM) and the recently-proposed important model, Minimax Probability Machine (MPM) consider data only either locally or globally. This new model is theoretically important in the sense that SVM and MPM can both be considered as its special case. Furthermore, the optimization of M 4 can be cast as a sequential conic programming problem, which can be solved efficiently. We describe the M 4 model definition, provide a clear geometrical interpretation, present theoretical justifications, propose efficient solving methods, and perform a series of evaluations on both synthetic data sets and real world benchmark data sets. Its comparison with SVM and MPM also demonstrates the advantages of our new model.
['Kaizhu Huang', 'Haiqin Yang', 'Irwin King', 'Michael R. Lyu']
Learning large margin classifiers locally and globally
473,901
The Atari 2600 games supported in the Arcade Learning Environment [Bellemare et al., 2013] all feature a known initial (RAM) state and actions that have deterministic effects. Classical planners, however, cannot be used off-the-shelf as there is no compact PDDL-model of the games, and action effects and goals are not known a priori. Indeed, there are no explicit goals, and the planner must select actions on-line while interacting with a simulator that returns successor states and rewards. None of this precludes the use of blind lookahead algorithms for action selection like breadth-first search or Dijkstra's yet such methods are not effective over large state spaces. We thus turn to a different class of planning methods introduced recently that have been shown to be effective for solving large planning problems but which do not require prior knowledge of state transitions, costs (rewards) or goals. The empirical results over 54 Atari games show that the simplest such algorithm performs at the level of UCT, the state-of-the-art planning method in this domain, and suggest the potential of width-based methods for planning with simulators when factored, compact action models are not available.
['Nir Lipovetzky', 'Miquel Ramírez', 'Hector Geffner']
Classical planning with simulators: results on the Atari video games
769,388
The classical view of cortical information processing is that of a bottom-up process in a feedforward hierarchy. However, psychophysical, anatomical, and physiological evidence suggests that top-down effects play a crucial role in the processing of input stimuli. Not much is known about the neural mechanisms underlying these effects. Here we investigate a physiologically inspired model of two reciprocally connected cortical areas. Each area receives bottom-up as well as top-down information. This information is integrated by a mechanism that exploits recent findings on somato-dendritic interactions. (1) This results in a burst signal that is robust in the context of noise in bottom-up signals. (2) Investigating the influence of additional top-down information, priming-like effects on the processing of bottom-up input can be demonstrated. (3) In accordance with recent physiological findings, interareal coupling in low-frequency ranges is characteristically enhanced by top-down mechanisms. The proposed scheme combines a qualitative influence of top-down directed signals on the temporal dynamics of neuronal activity with a limited effect on the mean firing rate of the targeted neurons. As it gives an account of the system properties on the cellular level, it is possible to derive several experimentally testable predictions.
['Markus Siegel', 'Konrad P. Kording', 'Peter König']
Integrating Top-Down and Bottom-Up Sensory Processing by Somato-Dendritic Interactions
337,371
Challenges of building a CMC corpus for analyzing writer's style by age: The DiDi project.
['Aivars Glaznieks', 'Egon Stemle']
Challenges of building a CMC corpus for analyzing writer's style by age: The DiDi project.
773,851
INTRODUCTION: SPECIAL ISSUE ON KNOWLEDGE REPRESENTATION AND ONTOLOGY RESEARCH
['Thomas Andreas Meyer', 'Mehmet A. Orgun']
INTRODUCTION: SPECIAL ISSUE ON KNOWLEDGE REPRESENTATION AND ONTOLOGY RESEARCH
483,612
Image Quality Assessment (IQA) plays an important role in assessing any new hardware, software, image acquisition techniques, image reconstruction or post-processing algorithms, etc. In the past decade, there have been various IQA methods designed to evaluate natural images. Some were used for the medical images but the use was limited. This paper reviews the recent advancement on IQA for medical images, mainly for Magnetic Resonance Imaging (MRI), Computed Tomography (CT), and ultrasonic imaging. Thus far, there is no gold standard of IQA for medical images due to various difficulties in designing a suitable IQA for medical images, and there are many different image characteristics and contents across various imaging modalities. No reference-IQA (NR-IQA) is recommended for assessing medical images because there is no perfect reference image in the real world medical imaging. We will discuss and comment on some useful and interesting IQA methods, and then suggest several important factors to be considered in designing a new IQA method for medical images. There is still great potential for research in this area.
['Li Sze Chow', 'Raveendran Paramesran']
Review of medical image quality assessment
689,714
Recently, it has been shown that prior to surgery a transrectal ultrasound (TRUS) study of the prostate and pubic arch can effectively determine pubic arch interference (PAI), a major stumbling block for the prostate brachytherapy (radioactive seed implantation) procedure. This PAI determination is currently being done with digital images taken directly from an ultrasound (US) machine. However, 70-75% of US machines used in prostate brachytherapy do not have a method to save or transfer digital image data for external use. To allow PAI assessment regardless of US platform and to keep costs to a minimum, we need to digitize the images from the US video output when there is no direct digital transfer capability. D/A and A/D conversions can introduce quantization error and other noises in these digitized images. The purpose of this work is to assess the image degradation caused by digitization and quantitatively evaluate whether after digitization it is still possible to accurately assess PAI. We used a PAI assessment algorithm (developed in previous research by our group) to predict the location of the pubic arch on both digital images and those captured after digitization. These predicted arch locations were compared to the "true" position of the pubic arch as established during surgery. Despite apparent image degradation due to the D/A and A/D conversions, we found no statistically significant difference between the accuracy of the predicted arch locations from the digitized images and those from the digital images. By demonstrating equally accurate determination of pubic arch locations using digital and digitized images, we conclude that TRUS-based PAI assessment can be easily and inexpensively performed in clinics where it is needed.
['Ken Haberman', 'Sayan D. Pathak', 'Yongmin Kim']
Effects of video digitization in pubic arch interference assessment for prostate brachytherapy
107,176
Virtual machines can greatly simplify grid computing by providing an isolated, well-known environment, while increasing security. Also, they can be used as the base technology to dynamically modify the computing elements of a grid, so providing an adaptive environment. In this paper we present a Grid architecture that allows to dynamically adapt the underlying hardware infrastructure to changing Virtual Organization (VO) demands. The backend of the system is able to provide on-demand virtual worker nodes to existing clusters and integrate them in any Globus-based Grid. In this way, we establish a basis to deploy self-adaptive Grids, which can support different VOs in shared physical infrastructures and dynamically adapt its software configuration. Experimental results on a prototyped testbed show less than a 10% overall performance loss including the hypervisor overhead.
['Manuel Donado Rodríguez', 'Daniel Tapiador', 'Javier Fontán', 'Eduardo Huedo', 'Rubén S. Montero', 'Ignacio Martín Llorente']
Dynamic Provisioning of Virtual Clusters for Grid Computing
365,525
Some Rules for Introducing Indexing Paths in a Primary File
['Michael Hatzopoulos', 'John G. Kollias']
Some Rules for Introducing Indexing Paths in a Primary File
125,601
The paper presents an application of satellite communication dedicated to medical assistance of people living in hazardous social-environmental conditions. This objective has been fulfilled by a consortium between telecommunication firms and public research bodies, in co-operation with the Italian Defense Staff, by means of a telemedicine network, involving hospitals in Sarajevo, Bosnia and Herzegovina, Albania, Romania and Italy. The network is a DAMA/PAMA (demand/permanent assignment multiple access) satellite network with VSAT class Earth stations, while telemedicine services are performed by means of a videoconference system that integrates satellite and terrestrial communication. A future development will be the creation of a telemedicine operating center, operative round the clock, for the promotion, maintenance and operative direction of the telemedicine services.
['Pietro Arcidiacono', 'Paolo Capodieci', 'Alessandro Garibbo', 'Claudio Giarrizzo', 'Romolo Grasso', 'Annamaria Raviola']
DESNET: a SCPC-DAMA network in satellite telemedicine applications
154,130
Survivable Grouped Routing Optical Networks with Dedicated Path Protection
['Hiroshi Hasegawa', 'Yojiro Mori', 'Ken-ichi Sato']
Survivable Grouped Routing Optical Networks with Dedicated Path Protection
832,003
This paper describes a new parallel raster terrain visibility (or view- shed) algorithm, based on the sweep-line model of (Van Kreveld 1996). Com- puting the terrain visible from a given observer is required for many GIS appli- cations, with applications ranging from radio tower siting to aesthetics. Pro- cessing the newly available higher resolution terrain data requires faster archi- tectures and algorithms. Since the main improvements on modern processors come from multi-core architectures, parallel programming provides a promis- ing means for developing faster algorithms. Our algorithm uses the economical and widely available shared memory model with OpenMP. Experimentally, our parallel speedup is almost linear. On 16 parallel processors, our algorithm is up to 12 times faster than the serial implementation.
['Chaulio R. Ferreira', 'Marcus V. A. Andrade', 'Salles V. G. Magalhães', 'W. Randolph Franklin', 'Guilherme C. Pena']
A Parallel Sweep Line Algorithm for Visibility Computation
673,160
The downlink channel covariance matrix (DCCM) is of vital importance in determining downlink beamforming weights for base station (BS) antenna array systems. For the frequency-division-duplex (FDD) mode, DCCM is difficult to obtain due to a lack of direct measurement of downlink channel responses. In this paper, a novel technique is proposed for estimating DCCM using uplink channel responses only, which does not need direction-of-arrival (DOA) estimation and its association. The downlink beamforming scheme is then proposed for wireless DS-CDMA systems, using the obtained DCCM information together with the so-called virtual uplink beamforming and power control technique. Computer simulations show that using the BS antenna array together with this new beamforming technique can provide larger system capacity than traditional DOA-based approaches, which just direct the main beam toward the desired user.
['Ying-Chang Liang', 'Francois P. S. Chin']
Downlink channel covariance matrix (DCCM) estimation and its applications in wireless DS-CDMA systems
68,719
This paper describes methods of designing large Arithmetic and Logical Units (ALUs) using multiple Programmable Logic Array (PLA) macros in which the outputs are obtained in one cycle corresponding to one pass through any PLA. The design is based on the well-known technique of providing conditional sums and group carries in parallel and selecting the proper sum using gating circuits. The PLA for each group of bits uses an adder design published by Weinberger in which each bit of the sum is formed from the EXCLUSIVE-OR of two outputs of the OR array. By placing the gating circuits in front of the EXCLUSIVE-OR circuits, the sums can be obtained using two array outputs for each bit and one additional OR array output for each internal string of bits. Also discussed are how ALUs containing more than two groups can obtain the group carries using a separate carry-look-ahead PLA macro and how this macro can be compressed by using special decoders and special physical design layout techniques. Additionally, the paper demonstrates how the PLAs can be used to provide detection of overflow and of zero results, and to also provide Boolean operations.
['Martin S. Schmookler']
Design of large ALUs using multiple PLA macros
196,375
This paper proposes a method to detect unknown words during natural reading of non-native language text by using eye-tracking features. A previous approach utilizes gaze duration and word rarity features to perform this detection. However, while this system can be used by trained users, its performance is not sufficient during natural reading by untrained users. In this paper, we 1) apply support vector machines (SVM) with novel eye movement features that were not considered in the previous work and 2) examine the effect of personalization. The experimental results demonstrate that learning using SVMs and proposed eye movement features improves detection performance as measured by F-measure and that personalization further improves results.
['Rui Hiraoka', 'Hiroki Tanaka', 'Sakriani Sakti', 'Graham Neubig', 'Satoshi Nakamura']
Personalized unknown word detection in non-native language reading using eye gaze
933,651
Statistical parameters of Ivan Franko's novel Perekhresni stežky (The Cross-Paths).
['Solomija N. Buk', 'Andrij Rovenchak']
Statistical parameters of Ivan Franko's novel Perekhresni stežky (The Cross-Paths).
742,936
In various application fields, high-speed cameras are used to analyze high-speed phenomena. Coded structured light projection methods have been proposed for acquiring three-dimensional images. Most of them are not suitable for measuring high-speed phenomena because the errors are caused when the measured objects move due to light projection of multiple patterns. In this paper, we propose a new coded structured light projection method that can select not only a temporal encoding but also a spatial encoding adaptively for obtaining three-dimensional images at a high-speed frame rate. We also develop an evaluation system that uses a DLP projector and an off-line high-speed video camera, and verify the effectiveness of the proposed method by showing the obtained three-dimensional shapes for moving objects.
['Idaku Ishii', 'Kenkichi Yamamoto', 'Kensuke Doi', 'Tokuo Tsuji']
High-speed 3D image acquisition using coded structured light projection
429,406
Exploiting the redundancy in election records to conduct useful audits and improve the system design process.
['Douglas W. Jones']
Auditing elections
712,432
In this paper, a self-learning tool for children belonging to age group 2-4 years is proposed where the subject can learn different educational content along with exercise. Our system is inspired by the traditional way to teach different elementary educational contents for example alphabets and digits, where the subject imitates the guide's movements and listen to the guide's voice to recognize alphabets or digits. Similarly the children learn by themselves several game rules and steps of games with spontaneous interest. A prototype of our proposed system is implemented, in which a subject can imitate the pose and instruction set of the virtual avatar (form) shown on the screen and the movement of the subject is recorded and analyzed in reverse using depth sensor. Depending on the analyzed data an audio visual feedback is given which is induced with game pleasure and reward function. If the participant can reproduce correct postures displayed by the avatar during training, the relevant reward function is displayed as an educational content. The children along with their guardian agreed that the system is interesting and can motivate the subject to do exercise along with learning i.e. Exer-learning. This system will also help the children having developmental delay to learn educational content in significant way.
['Sonia Nandi', 'Suman Deb', 'Mitali Sinha']
Augmented Exer-Learning Tool Using Ultrasonic Depth Visualization of Movement
989,169
In many practical applications of supervised learning the task involves the prediction of multiple target variables from a common set of input variables. When the prediction targets are binary the task is called multi-label classification, while when the targets are continuous the task is called multi-target regression. In both tasks, target variables often exhibit statistical dependencies and exploiting them in order to improve predictive accuracy is a core challenge. A family of multi-label classification methods address this challenge by building a separate model for each target on an expanded input space where other targets are treated as additional input variables. Despite the success of these methods in the multi-label classification domain, their applicability and effectiveness in multi-target regression has not been studied until now. In this paper, we introduce two new methods for multi-target regression, called stacked single-target and ensemble of regressor chains, by adapting two popular multi-label classification methods of this family. Furthermore, we highlight an inherent problem of these methods--a discrepancy of the values of the additional input variables between training and prediction--and develop extensions that use out-of-sample estimates of the target variables during training in order to tackle this problem. The results of an extensive experimental evaluation carried out on a large and diverse collection of datasets show that, when the discrepancy is appropriately mitigated, the proposed methods attain consistent improvements over the independent regressions baseline. Moreover, two versions of Ensemble of Regression Chains perform significantly better than four state-of-the-art methods including regularization-based multi-task learning methods and a multi-objective random forest approach.
['Eleftherios Spyromitros-Xioufis', 'Grigorios Tsoumakas', 'William Groves', 'Ioannis P. Vlahavas']
Multi-target regression via input space expansion: treating targets as inputs
644,606
This paper presents a high-level approach for assessing the performance behavior of complex scientific applications running on a high-performance system through simulation. The proposed methodology relies on high-level descriptions of both application and system. The application is described in MetaPL, an XML-based description language, and the system is modeled and simulated by using HeSSE, an extensible distributed heterogeneous system (DHS) simulator. This modeling technique is applied to the performance analysis of a real-world scientific application (LAPWO), running on a cluster of SMP nodes. Extendibility features of both MetaPL and HeSSE were largely used, developing extensions for the MetaPL language and new components for the simulator. The paper closes with the validation of the performance model, obtained through the comparison of the predicted performance results with measurements on test runs of the application on the real system.
['Thomas Fahringer', 'Nicola Mazzocca', 'Massimiliano Rak', 'Sabri Pllana', 'Umberto Villano', 'Georg Madsen']
Performance modeling of scientific applications: scalability analysis of LAPW0
176,850
In this paper, the problem of capacity analysis of data-hiding techniques in a game information-theoretic framework is considered. Capacity is determined by the stochastic model of the host image, by the distortion constraints, and by the side information about the watermarking channel state available at the encoder and at the decoder. The importance of the proper modeling of image statistics is emphasized, and for this purpose, a novel stochastic nonstationary image model is proposed that is based on geometrical priors, the so-called edge process model. Being mathematically simple and tractable, the edge process model outperforms the estimation-quantization (EQ) and spike process models in reference applications such as denoising. Finally, this model allows us to obtain a realistic estimate of maximal embedding rates, and in particular, it is shown that the expected capacity limit of real images is significantly lower than previously reported.
['Sviatoslav Voloshynovskiy', 'Oleksiy J. Koval', 'Mehmet Kivanc Mihcak', 'Thierry Pun']
The edge process model and its application to information-hiding capacity analysis
513,284
Based on the OFDM signal reconstructed by data after decision and channel estimation, a new OFDM carriers frequency offset blind estimation algorithm is proposed. This method can be used to OFDM carrier frequency tracking. This estimator has advantages such as implementation facility, low computations, which only need one IFFT. Simulation results show, this tracking method has very high precision.
['Yan Du', 'Xiaoinin Zhang', 'Pengcheng Zhu']
A signal reconstruction CFO blind tracking algorithm for OFDM systems in multipath channels
321,491
The rapid growth in the volume of many real-world graphs (e.g., social networks, web graphs, and spatial networks) has led to the development of various vertex-centric distributed graph computing systems in recent years. However, real-world graphs from different domains have very different characteristics, which often create bottlenecks in vertex-centric parallel graph computation. We identify three such important characteristics from a wide spectrum of real-world graphs, namely (1)skewed degree distribution, (2)large diameter, and (3)(relatively) high density. Among them, only (1) has been studied by existing systems, but many real-world power-law graphs also exhibit the characteristics of (2) and (3). In this paper, we propose a block-centric framework, called Blogel, which naturally handles all the three adverse graph characteristics. Blogel programmers may think like a block and develop efficient algorithms for various graph problems. We propose parallel algorithms to partition an arbitrary graph into blocks efficiently, and block-centric programs are then run over these blocks. Our experiments on large real-world graphs verified that Blogel is able to achieve orders of magnitude performance improvements over the state-of-the-art distributed graph computing systems.
['Da Yan', 'James Cheng', 'Yi Lu', 'Wilfred Ng']
Blogel: a block-centric framework for distributed computation on real-world graphs
622,831
In this paper we consider the task of locating salient group-structured features in potentially high-dimensional images; the salient feature detection here is modeled as a Robust Principal Component Analysis problem, in which the aim is to locate groups of outlier columns embedded in an otherwise low rank matrix. We adapt an adaptive compressive sensing method from our own previous work (which examined the task of identifying arbitrary sets of outlier columns in large matrices) to settings where the outlier columns occur in groups, and establish theoretical results certifying that accurate group-structured inference is achievable using very few linear measurements of the image, subject to some (arguably) minor structural assumptions on the image itself. We also demonstrate, through extensive numerical simulations, our proposed algorithm in a salient object detection task, and show that it simultaneously achieves low sample and computational complexity, while exhibiting performance comparable to state-of-the-art methods that acquire and process the entire image.
['Xingguo Li', 'J. Haupt']
Locating salient group-structured image features via adaptive compressive sensing
655,916
Design of Multiple-Valued Logic Circuits Using Graph-Based Evolutionary Synthesis.
['Masanori Natsui', 'Naofumi Homma', 'Takafumi Aoki', 'Tatsuo Higuchi']
Design of Multiple-Valued Logic Circuits Using Graph-Based Evolutionary Synthesis.
736,885
A potential theoretical method for exploring the engineering first year retention problem
['Craig A. Watterson', 'Dale A. Carnegie', 'Marc Wilson', 'Bernadette Knewstubb']
A potential theoretical method for exploring the engineering first year retention problem
912,543
In order to handle a material with either a delicate surface or an air permeable structure, a novel nozzle was designed and developed. This nozzle utilises the phenomena of the radial air outflow. It is envisaged that this new nozzle will handle materials by eliminating surface marking and air permeable structure problems.
['Babur Ozcelik', 'Fehmi Erzincanli']
A non-contact end-effector for the handling of garments
527,887
Virtual machine (VM) consolidation is a promising approach for improving energy efficiency of the datacenter by increasing the resource utilization of physical machines. However, the live migration technology that VM consolidation relies on is costly in itself, and this migration cost is usually heterogeneous as well as the datacenter. This paper focuses on how to pay limited migration costs to save as much energy as possible via VM consolidation in a heterogeneous cloud environment. That is, how to minimize the energy consumption while keeping most of the VMs in the datacenter unmoved. To capture these two conflicting objectives, a migration cost estimation method is first proposed and then a consolidation score function is defined for overall evaluation. To maximize the consolidation score, an improved grouping genetic algorithm (IGGA) based on a greedy heuristic and a swap operation is proposed for VM consolidation. Experiments show that IGGA performs better than existing consolidation methods.
['Quanwang Wu', 'Fuyuki Ishikawa']
Heterogeneous Virtual Machine Consolidation Using an Improved Grouping Genetic Algorithm
548,378
Implementación sobre FPGA de la estrategia evolutiva CMA-ES para optimización numérica
['Leopoldo Urbina', 'Carlos A. Duchanoy', 'Marco A. Moreno-Armendáriz', 'Derlis Lara', 'Hiram Calvo']
Implementación sobre FPGA de la estrategia evolutiva CMA-ES para optimización numérica
726,728
This paper presents a large scale dataset of vision (stereo and RGB-D), laser and proprioceptive data collected over an extended duration by a Willow Garage PR2 robot in the 10 story MIT Stata Center. As of September 2012 the dataset comprises over 2.3 TB, 38 h and 42 km (the length of a marathon). The dataset is of particular interest to robotics and computer vision researchers interested in long-term autonomy. It is expected to be useful in a variety of research areas-robotic mapping (long-term, visual, RGB-D or laser), change detection in indoor environments, human pattern analysis, long-term path planning. For ease of use the original ROS 'bag' log files are provided and also a derivative version combining human readable data and imagery in standard formats. Of particular importance, this dataset also includes ground-truth position estimates of the robot at every instance (to typical accuracy of 2 cm) using as-built floor-plans-which were carefully extracted using our software tools. The provision of ground-truth for such a large dataset enables more meaningful comparison between algorithms than has previously been possible.
['Maurice F. Fallon', 'Hordur Johannsson', 'Michael Kaess', 'John J. Leonard']
The MIT Stata Center dataset
221,767
A single-phase half-bridge switching mode rectifier is presented to draw a sinusoidal line current, to achieve power factor correction and to maintain the dc-link voltage constant. Four active switches with voltage stress of half dc bus voltage are used in the proposed rectifier to generate a unipolar PWM voltage waveform on the ac terminal voltage. There is no clamping diode in the proposed rectifier compared with the neutral point clamped converter to achieve three-level PWM operation. Two control loops are used in the proposed control scheme. In the outer control loop, a proportional integral voltage controller is used to regulate the dc-link voltage. A phase lock loop circuit is adopted to generate a sinusoidal waveform in phase with mains voltage to achieve power factor correction. In the inner control loop, a carrier-based current controller is used to track the line current command. The experimental results are presented to verify the effectiveness of the proposed control algorithm.
['Bor-Ren Lin', 'Tsung-Yu Yang']
Single-phase three-level converter for power factor correction
365,962
Due to the explosive growth of social network service resulting from the popularity of smart devices, online relations and activities are now affecting the behaviors of many people. On that account, the interest and importance of social network activities on Internet continue to grow. This study defines the social data with the following four factors: object, user, direction and distance. Moreover, this study quantifies the structured data such as number of responses and number of friends and the unstructured data such as difference between cause time and response time, preference and response type in relation to the object based activities of social network service (SNS) users in terms of time axis. This study then proposes the model to measure the influence direction and influence strength (or distance). In addition, this study models and explains the process regarding the system to collect and analyze the data for influence measurement and also the influence measurement technique using the sample data collected on facebook.
['Seoung-hyun Koh', 'Yen-Yoo You', 'Do-Sung Na']
Object-based dynamic influence measurement model (DIMM) using social data (on facebook)
932,340
A perturbation estimator using the theory of variable structure systems is proposed to enhance the robustness of a pole-placement controller design. In its ideal form, the pole-placement design using feedback-linearization technique achieves a desired performance in nonlinear time-varying systems. However, its performance deteriorates rapidly with the presence of disturbance and parametric uncertainties, referred to as perturbation. The estimate generated by the proposed perturbation estimator is incorporated as an additional input to rectify the uncertainties in the nominal control model of the pole-placement design. The proposed scheme requires neither the measurement of the time derivative of the state vector nor the precise knowledge of system parameters, hut rather the bounds on system perturbation. Chatter and the adverse effects of conservative bounds on system perturbation, often encountered in conventional sliding-mode control (SMC), are alleviated for the controlled plant by the proposed scheme. The benefits of this scheme are demonstrated in this study practically on a magnetic levitation system and its performance is compared with that of the conventional SMC scheme. >
['Yu Sheng Lu', 'Jian Shiang Chen']
Design of a perturbation estimator using the theory of variable-structure systems and its application to magnetic levitation systems
145,974
Acetylcholine (ACh) plays an important role in a wide variety of cognitive tasks, such as perception, selective attention, associative learning, and memory. Extensive experimental and theoretical work in tasks involving learning and memory has suggested that ACh reports on unfamiliarity and controls plasticity and effective network connectivity. Based on these computational and implementational insights, we develop a theory of cholinergic modulation in perceptual inference. We propose that ACh levels reflect the uncertainty associated with top-down information, and have the effect of modulating the interaction between top-down and bottom-up processing in determining the appropriate neural representations for inputs. We illustrate our proposal by means of an hierarchical hidden Markov model, showing that cholinergic modulation of contextual information leads to appropriate perceptual inference.
['Angela J. Yu', 'Peter Dayan']
Acetylcholine in cortical inference
152,610
A mathematical model of the calcium transient in urinary bladder smooth muscle cells
['Vijay Dave', 'Chitaranjan Mahapatra', 'Rohit Manchanda']
A mathematical model of the calcium transient in urinary bladder smooth muscle cells
677,473
The role of physical constraints in natural and artifical manipulation
['Pietro Morasso', 'Ferdinando A. Mussa-Ivaldi']
The role of physical constraints in natural and artifical manipulation
285,251
Convolutional neural networks (CNNs) are revolutionizing a variety of machine learning tasks, but they present significant computational challenges. Recently, FPGA-based accelerators have been proposed to improve the speed and efficiency of CNNs. Current approaches construct a single processor that computes the CNN layers one at a time; this single processor is optimized to maximize the overall throughput at which the collection of layers are computed. However, this approach leads to inefficient designs because the same processor structure is used to compute CNN layers of radically varying dimensions. #R##N#We present a new CNN accelerator paradigm and an accompanying automated design methodology that partitions the available FPGA resources into multiple processors, each of which is tailored for a different subset of the CNN convolutional layers. Using the same FPGA resources as a single large processor, multiple smaller specialized processors result in increased computational efficiency and lead to a higher overall throughput. Our design methodology achieves 1.51x higher throughput than the state of the art approach on evaluating the popular AlexNet CNN on a Xilinx Virtex-7 FPGA. Our projections indicate that the benefit of our approach increases with the amount of available FPGA resources, already growing to over 3x over the state of the art within the next generation of FPGAs.
['Yongming Shen', 'Michael Ferdman', 'Peter A. Milder']
Maximizing CNN Accelerator Efficiency Through Resource Partitioning
829,376
Pervasive technologies are enabling an increasingly data-rich world that is mediated through a broad spectrum of often highly interdependent systems. The data science surrounding these systems is rapidly transforming nearly every aspect of our lives. But how trustworthy are the systems and data upon which we have come to rely? This article explores the complex collaborations and interdependencies that mediate trust-formation and examines six challenges in generating and sustaining trust in the context of data-rich systems.
['Brandin Knowles']
Emerging Trust Implications of Data-Rich Systems
924,615
Finding the accurate position of an eye is crucial for mobile iris recognition system in order to extract the iris region quickly and correctly. Unfortunately, this is very difficult to accomplish when a person is wearing eyeglasses because of the interference from the eyeglasses. This paper proposes an eye detection method that is robust to eyeglass interference in mobile environment. The proposed method comprises two stages: eye candidate generation and eye validation. In the eye candidate generation stage, multi-scale window masks consisting of 2 × 3 subblocks are used to generate all image blocks possibly containing an eye image. In the ensuing eye validation stage, two methods are employed to determine which blocks actually contain true eye images and locate their precise positions as well: the first method searches for the glint of an NIR illuminator on the pupil region. If this first method fails, the next method computes the intensity difference between the assumed pupil and its surrounding region using multi-scale 3 × 3 window masks. Experimental results show that the proposed method detects the eye position more accurately and quickly than competing methods in the presence of interference from eyeglass frames.
['Yujin Jung', 'Dongik Kim', 'Byung-Jun Son', 'Jaihie Kim']
An eye detection method robust to eyeglasses for mobile iris recognition
893,100
New standards and initiatives in the U.S. electric power grid are moving in the direction of a smarter grid. Media attention has focused prominently on smart meters in distribution systems, but big changes are also occurring in the domains of protection, control, and Supervisory Control and Data Acquisition (SCADA) systems. These changes promise to enhance the reliability of the electric power grid and to allow it to safely operate closer to its limits, but there is also a real danger concerning the introduction of network communication vulnerabilities to so-called cyber attacks. This article advocates the use of a reputation-based trust management system as one method to mitigate such attacks. A simulated demonstration of the potential for such systems is illustrated in the domain of backup protection systems. The simulation results show the promise of this proposed technique.
['Jose E. Fadul', 'Kenneth M. Hopkinson', 'Christopher Sheffield', 'James T. Moore', 'Todd R. Andel']
Trust Management and Security in the Future Communication-Based "Smart" Electric Power Grid
166,943
Optically Driven Micromanipulators with Rotating Arms.
['Shoji Maruo', 'Yojiro Hiratsuka']
Optically Driven Micromanipulators with Rotating Arms.
993,846
We prove a conjecture of Cohn and Propp, which refines a conjecture of Bosley and Fidkowski about the symmetry of the set of alternating sign matrices (ASMs). We examine data arising from the representation of an ASM as a collection of paths connecting $2n$ vertices and show it to be invariant under the dihedral group $D_{2n}$ rearranging those vertices, which is much bigger than the group of symmetries of the square. We also generalize conjectures of Propp and Wilson relating some of this data for different values of $n$.
['Benjamin Wieland']
A Large Dihedral Symmetry of the Set of Alternating Sign Matrices
177,100
(STBC) for MIMO Gaussian wiretap channels. It is assumed that the transmitter has the receiver's channel state information, but not that of the eavesdropper. We first propose a full-rate STBC that provides separate decoding complexity (rather than pairwise) at the intended receiver, while requiring an exhaustive search for Maximum Likelihood (ML) decoding at the eavesdropper. Next we make the code more secure by including artificial noise symbols which (for the asymptotically high SNR regime) are aligned with each other and subtracted from the information symbols at the intended receiver, but which can not be cancelled at the eavesdropper. Simulations demonstrate the enhanced physical-layer security that results.
['S. Ali. A. Fakoorian', 'Hamid Jafarkhani', 'A. Lee Swindlehurst']
Secure space-time block coding via artificial noise alignment
936,398
Crossing Healthgrid borders: early results in medical imaging.
['Sílvia Delgado Olabarriaga', 'Tristan Glatard', 'Andreas Hoheisel', 'Aart J. Nederveen', 'Dagmar Krefting']
Crossing Healthgrid borders: early results in medical imaging.
753,359
Reverse Logistics for Solid Waste is an important problem concerning about economic and environment. This paper aims at determine the optimal locations and numbers of treatment facilities and transfer stations for municipal solid wastes. The proposed method is the combination of a fuzzy multi-objective programming and a genetic algorithm. The original fuzzy multiple objectives are appropriately converted to a single unified "min- max" goal, which makes it easy to design a genetic algorithm for the problem solving. So, a reverse logistics network for municipal solid waste is constructed. Finally, a case study of China proves the availability of our approach.
['Bo He', 'Chao Yang', 'Mingming Ren']
A Fuzzy Multi-objective Programming for Optimization of Reverse Logistics for Solid Waste through Genetic Algorithms
507,988
Suffix trees and suffix arrays are fundamental full-text index data structures to solve problems occurring in string processing. Since suffix trees and suffix arrays have different capabilities, some problems are solved more efficiently using suffix trees and others are solved more efficiently using suffix arrays. We consider efficient index data structures with the capabilities of both suffix trees and suffix arrays without requiring much space. When the size of an alphabet is small, enhanced suffix arrays are such index data structures. However, when the size of an alphabet is large, enhanced suffix arrays lose the power of suffix trees. Pattern searching in an enhanced suffix array takes O(m|Σ|) time while pattern searching in a suffix tree takes O(mlog |Σ|) time where m is the length of a pattern and Σ is an alphabet.#R##N##R##N#In this paper, we present linearized suffix trees which are efficient index data structures with the capabilities of both suffix trees and suffix arrays even when the size of an alphabet is large. A linearized suffix tree has all the functionalities of the enhanced suffix array and supports the pattern search in O(mlog |Σ|) time. In a different point of view, it can be considered a practical implementation of the suffix tree supporting O(mlog |Σ|)-time pattern search.#R##N##R##N#In addition, we also present two efficient algorithms for computing suffix links on the enhanced suffix array and the linearized suffix tree. These are the first algorithms that run in O(n) time without using the range minima query. Our experimental results show that our algorithms are faster than the previous algorithms.
['Dong Kyue Kim', 'Minhwan Kim', 'Heejin Park']
Linearized Suffix Tree: an Efficient Index Data Structure with the Capabilities of Suffix Trees and Suffix Arrays
449,902
A general hill-climbing attack to biometric systems based on a modification of the downhill simplex algorithm is presented. The scores provided by the matcher are used in this approach to adapt iteratively an initial estimate of the attacked template to the specificities of the client being attacked. The proposed attack is evaluated on a competitive feature-based signature verification system over both the MCYT and the BiosecurID databases (comprising 330 and 400 users, respectively). The results show a very high efficiency of the hill-climbing algorithm, which successfully bypassed the system for over 90% of the attacks with a remarkably low number of scores needed.
['Marta Gomez-Barrero', 'Javier Galbally', 'Julian Fierrez', 'Javier Ortega-Garcia']
Hill-climbing attack based on the uphill simplex algorithm and its application to signature verification
291,450
Directed graphs encode meaningful dependencies among objects ubiquitously. This paper introduces new and simple representations for labeled directed graphs with the properties of being succinct (space is information-theoretically optimal); in which we avoid exploiting a-priori knowledge on digraph regularity such as triangularity, separability, planarity, symmetry and sparsity. Our results have direct implications to model directed graphs by using single integer numbers effectively, which is significant to enable canonical (generation of graph instances is unique) and efficient (coding and decoding take polynomial time) encodings for learning and optimization algorithms. To the best of our knowledge, the proposed representations are the first known in the literature.
['Victor Parque', 'Tomoyuki Miyashita']
On succinct representation of directed graphs
959,534
Utility computing is being gradually realized as exemplified by cloud computing. Outsourcing computing and storage to global-scale cloud providers benefits from high accessibility, flexibility, scalability, and cost-effectiveness. However, users are uneasy outsourcing the storage of sensitive data due to security concerns. We address this problem by presenting SeMiNAS---an efficient middleware system that allows files to be securely outsourced to providers and shared among geo-distributed offices. SeMiNAS achieves end-to-end data integrity and confidentiality with a highly efficient authenticated-encryption scheme. SeMiNAS leverages advanced NFSv4 features, including compound procedures and data-integrity extensions, to minimize extra network round trips caused by security meta-data. SeMiNAS also caches remote files locally to reduce accesses to providers over WANs. We designed, implemented, and evaluated SeMiNAS, which demonstrates a small performance penalty of less than 26% and an occasional performance boost of up to 19% for Filebench workloads.
['Ming Chen', 'Erez Zadok', 'Arun Olappamanna Vasudevan', 'Kelong Wang']
SeMiNAS: A Secure Middleware for Wide-Area Network-Attached Storage
813,927
Decision Analysis: A Personal Account of How It Got Started and Evolved
['Howard Raiffa']
Decision Analysis: A Personal Account of How It Got Started and Evolved
59,530
Many emerging communication technologies significantly increase the complexity of the physical layer and have dramatically increased the number of operating configurations. To ensure maximum performance, designers have to optimize their algorithm implementations, which requires for comprehensive performance testing in all possible operating modes various channel conditions. This paper presents a flexible and affordable framework for baseband algorithm development and performance verification for digital communication systems with an arbitrary number of modules, each operating at a possibly different sampling rate with various latencies. The proposed architecture is scalable to support complex scenarios, such as multiple antenna systems, and is compact enough to be implemented within a single field-programmable gate array.
['Amirhossein Alimohammad', 'Saeed Fouladi Fard', 'Bruce F. Cockburn']
A flexible layered architecture for accurate digital baseband algorithm development and verification
220,870
Human tactile perception incorporates two important characteristics - self-reference and bidirectionality. Self-reference is an important factor when evaluating tactile sensations. Bidirectionality contributes to sensing capability enhancement. In this paper, a tactile sensor including self-reference and bidirectionality is presented. The sensor with two microphones is mounted on a human finger. Users can apply the sensor while retaining their normal tactile perceptions and simultaneously obtaining sound and skin deformation data based on the mechanical interaction between the finger and the object. Experimental results on roughness evaluations and convexity detections show the validity of the proposed sensing method and its potential for use in various applications.
['Yoshihiro Tanaka', 'Yoshihiro Horita', 'Akihito Sano', 'Hideo Fujimoto']
Tactile sensing utilizing human tactile perception
359,903
We describe the main characteristics of version 0.61 of the E equational theorem prover. E is based on superposition (with literal selection) and rewriting. A particular strength of E is the ability to control the proof search very well. This is reflected by a very powerful and flexible interface for the specification of clause selection functions, and by a wide variety of functions for the selection of inference literals. We discuss some important aspects of the implementation and demonstrate the performance of the prover by presenting experimental results on the TPTP. Finally, we describe our future plans for the system.
['Stephan Schulz']
System Abstract: E 0.61
188,946
A Framework for Improving the Quality of Operation in a Virtual Enterprise
['P. M. Wognum', 'Edward Faber']
A Framework for Improving the Quality of Operation in a Virtual Enterprise
21,902
The recent surge in popularity of crowdsourcing has brought with it a new opportunity for engaging human intelligence in the process of data analysis. Crowdsourcing provides a fundamental mechanism for enabling online workers to participate in tasks that are either too difficult to be solved solely by a computer or too expensive to employ experts to perform. In the field of social science, four elements are required to form a wise crowd - Diversity of Opinion, Independence, Decentralization and Aggregation. However, while the other three elements are already studied and implemented in current crowdsourcing platforms, the 'Diversity of Opinion' has not been functionally enabled. In this paper, we address the algorithmic optimizations towards the diversity of opinion of crowdsourcing marketplaces.#R##N##R##N#From a computational perspective, in order to build a wise crowd, we need to quantitatively modeling the diversity, and take it into consideration for constructing the crowd. In a crowdsourcing marketplace, we usually encounter two basic paradigms for worker selection: building a crowd to wait for tasks to come and selecting workers for a given task. Therefore, we propose our Similarity-driven Model (S-Model) and Task-driven Model (T-Model) for both of the paradigms. Under both of the models, we propose efficient and effective algorithms to enlist a budgeted number of workers, which have the optimal diversity. We have verified our solutions with extensive experiments on both synthetic datasets and real data sets.
['Ting Wu', 'Lei Chen', 'Pan Hui', 'Chen Jason Zhang', 'Weikai Li']
Hear the whole story: towards the diversity of opinion in crowdsourcing markets
601,058
A distributed detection method is proposed to detect single stage multi-point (SSMP) attacks on a Cyber Physical System (CPS). Such attacks aim at compromising two or more sensors or actuators at any one stage of a CPS and could totally compromise a controller and prevent it from detecting the attack. However, as demonstrated in this work, using the flow properties of water from one stage to the other, a neighboring controller was found effective in detecting such attacks. The method is based on physical invariants derived for each stage of the CPS from its design. The attack detection effectiveness of the method was evaluated experimentally against an operational water treatment testbed containing 42 sensors and actuators. Results from the experiments point to high effectiveness of the method in detecting a variety of SSMP attacks but also point to its limitations. Distributing the attack detection code among various controllers adds to the scalability of the proposed method.
['Sridhar Adepu', 'Aditya P. Mathur']
Distributed Detection of Single-Stage Multipoint Cyber Attacks in a Water Treatment Plant
755,654
Specification of monitored context properties and their influence on behavior of Web services and management activities is a prerequisite for contextsensitive operation, which is a characteristic management issue for mobile/embedded Web services. Our Web Service Offerings Language (WSOL) 1.1 provided formal specification of classes of service, different types of constraints, and management statements for Web services. In this paper, we present new WSOL version 1.4 that enables specification of monitored context properties, context monitoring schedules, context management parties, and context exchange mechanisms. It also enables descriptions of how context influences behavior of Web services, monitoring of Web services, and dynamic adaptation of Web service compositions. We verified feasibility of these improvements using syntax checks of the WSOL 1.4 grammar and modifications of prototype WSOL-related tools. We validated them on emulated case studies. An important novelty of this work is using context information in contract-based management of Web services.
['Vladimir Tosic', 'Hanan Lutfiyya', 'Yazhe Tang']
Web Service Offerings Language (WSOL) Support for Context Management of Mobile/Embedded XML Web Services
351,332
this project is a proposal, based on innovative Bluetooth Marketing Services, whose main objective is to facilitate the dissemination of information via Bluetooth in tourist-commercial sectors. This proposal is supported by the analysis, design and construction of a software application resulting from the abstraction of previous analysis and design of a directory-based service.
['Octavio José Salcedo Parra', 'Samuel Sabogal', 'Lilia Castellanos']
A Based BlueTooth Tourist Service Prototype
465,625
There has been a growing demand for the development of tools to manage enterprise communication networks. A management information database is the heart of a network management system-it provides the interface between all functions of the network management system and, therefore, has to provide sophisticated functionality allied with high performance. The authors introduce the design of MANDATE (MAnaging Networks using DAtabase TEchnology), a proposed database system for effectively supporting the management of large enterprise networks. The MANDATE design makes a conscious attempt to take advantage of the special characteristics of network data and transactions, and of advances in database technology, to efficiently derive some of the required management functionality. >
['Jayant R. Haritsa', 'Michael O. Ball', 'Nick Roussopoulos', 'Anindya Datta', 'John S. Baras']
MANDATE: managing networks using database technology
225,464
The authors address the problem of transmission scheduling and buffer management at a switch in a high-speed packet-switched network. Future computer networks are expected to carry bursty real time multimedia traffic with specific time-delay requirements. We propose a technique to schedule competing packets at a switch in a packet-switched network with two objectives: (i) maximizing the number of packets that would reach their destination before their deadline has been reached; (ii) minimizing the number of packets dropped in transit in the network. The method is broadly based on the popular least-laxity heuristic in real time process scheduling and is combined with buffer management at the switch.
['Sudhir M. Rao', 'Albert Mo Kim Cheng']
Scheduling and routing of real-time multimedia traffic in packet-switched networks
508,633
There is a growing trend towards convergence of telecommunication and data networks in order to support a richer set of services and applications. At the same time, increasing diversity and density of network access technologies has made the goal of providing connectivity anytime and anywhere a real possibility. Another important development is the emergence of small, low-complexity user owned networks, such as personal area networks and body area networks. Dynamic interworking, also known as network composition, between networks of different types and sizes is essential in the push towards convergence, as well as to realize truly seamless connectivity between heterogeneous access networks. Dynamic interworking requires signalling between different elements of the control planes of the different networks in order to coordinate the control functions and resources of the networks concerned. In this paper, we present the generic ambient network signalling protocol suite to address the diverse signalling requirements for dynamic interworking of networks.
['Nadeem Akhtar', 'Rui Campos', 'Cornelia Kappler', 'Pekka Pääkkönen', 'Petteri Pöyhönen', 'Di Zhou']
GANS: A Signalling Framework for Dynamic Interworking Between Heterogeneous Networks
535,197
Approximation Algorithms for Facility Dispersion.
['Daniel J. Rosenkrantz', 'S. S. Ravi', 'Giri Kumar Tayi']
Approximation Algorithms for Facility Dispersion.
781,576
Micromanipulation by microrobots has become an issue of primary importance in industry and biomedicine, since human manual capabilities are restricted to certain tolerances. The manipulation of biological cells or the assembly of a microsystem composed of several microcomponents are good examples. An automated microrobot-based micromanipulation desktop station has been developed at the University of Karlsruhe. The process of assembly takes place in the field of view of a light optical microscope. This paper focuses on motion control problems of the piezo-driven microrobots employed by the station. The ability to adapt itself to the process requirements is of great importance for micromanipulation robots. They must be able to operate in a partially defined environment and to ensure reasonable behaviour in unpredicted situations. A neural control concept based on a reference model is proposed as a solution. It is shown, that the neural controller is able to learn the desired behaviour. It considerably outperforms an analytically designed linear controller in the real environment.
['Karoly Santa', 'Sergej Fatikow']
Development of a neural controller for motion control of a piezoelectric three-legged micromanipulation robot
530,188
We present a navigation system using multiple sensors for unknown and dynamic indoor environments. To achieve the robustness and flexibility of the mobile robot, we propose a new behavior-based architecture with three groups of clustered (reflexive, purposive, and adaptive) agents that realizes both efficiency in attaining the mission of the robot and robustness against the various kinds or failures that may occur in a dynamic environment. Basic behaviors required for navigation, such as, avoiding obstacles, moving towards free space, and following targets, are redundantly developed as agents and combined in the behavior-based system architecture.
['In So Kweon', 'Yoshinori Kuno', 'Mutsumi Watanabe', 'Kazunori Onoguchi']
Behavior-based Intelligent Robot In Dynamic Indoor Environments
348,276
An effective way to improve a classification method's performance is to create ensembles of classifiers. Two elements are believed to be important in constructing an ensemble: (a) the performance of each individual classifier; and (b) diversity among the classifiers. Nevertheless, most works based on diversity suggest that there exists only weak correlation between classifier performance and ensemble accuracy. We propose compound diversity functions which combine the diversities with the performance of each individual classifier, and show that there is a strong correlation between the proposed functions and ensemble accuracy. Calculation of the correlations with different ensemble creation methods, different problems and different classification algorithms on 0.624 million ensembles suggests that most compound diversity functions are better than traditional diversity measures. The population-based Genetic Algorithm was used to search for the best ensembles on a handwritten numerals recognition problem and to evaluate 42.24 million ensembles. The statistical results indicate that compound diversity functions perform better than traditional diversity measures, and are helpful in selecting the best ensembles.
['Albert Hung-Ren Ko', 'Robert Sabourin', 'J R Alceu De Souza Britto']
COMPOUND DIVERSITY FUNCTIONS FOR ENSEMBLE SELECTION
183,155
This paper presents a innovative algorithm to estimate the motion parameters of a mobile robot equipped with a radial laser rangefinder. Our method is based on the spatial and temporal linearization of the range function, which leads to a velocity constraint equation for the scanned points. The proposed formulation computes the motion vectors of the scanned points as they move from scan to scan in the sequence. This motion field can be very useful in a number of applications including detection and tracking of moving objects. Experimental results are presented, showing that good results are achieved with both real and synthetic data.
['Javier Gonzalez', 'Rafael M. Gutiérrez']
Mobile robot motion estimation from a range scan sequence
297,097
The trend of “open data” coupled with the recent advancement in web development technologies and the proliferation of JavaScript frameworks has helped popularize programming of interactive web applications. Still, some of the common features of today's web applications that access data from own data stores or from web services require a complex setup or a significant amount of programming knowledge, and thus make it hard for developers to quickly prototype applications and iterate on solutions. Therefore, we propose Endev, a declarative framework for prototyping applications that use cloud data storage or web service data. By not needing to write any JavaScript code or set up any servers, Endev provides a low learning threshold. We show that Endev is perceived useful and easy to use through a study with 15 developers.
['Filip Kis', 'Cristian Bogdan']
Declarative setup-free web application prototyping combining local and cloud datastores
943,354
Geographic Information Systems (GIS) hydrologic modeling techniques are used to better understand the surface-flow characteristics in the Prairie Pothole Region (PPR) of North America. This research uses an airborne Interferometric Synthetic Aperture Radar (IFSAR)-derived digital terrain model (DTM) as a base for developing a hydrologically-correct DEM and derivative products. The IFSAR DTM is assessed for accuracy and ability to resolve wetland features. A wetland mask is developed to selectively fill the DTM and from it products such as wetland catchments and drainage linkages are derived and interpreted. Study sites in the PPR are two surveyed and closely monitored wetland complexes, Crystal Springs and Orchid Meadows in Deuel County, South Dakota, USA.
['Janet H. Gritzner']
Modeling surface-flow characteristics in glaciated landscapes
223,685
Radial basis-function networks (RBFNs) have recently attracted interest, because of their advantages over multilayer perceptrons as they are universal approximators but achieve faster convergence since only one layer of weights is required. The least squares method is the most popularly used in estimating the synaptic weights which provides optimal results if the underlying error distribution is Gaussian. However, the generalization performance of the networks deteriorates for realistic noise whose distribution is either unknown or non-Gaussian; in particular, it becomes very bad if outliers are present. In this paper we propose a positive-breakdown learning algorithm for RBFNs by applying the breakdown point approach in robust regression such that any assumptions about or estimation of the error distribution are avoidable. The expense of losing efficiency in the presence of Gaussian noise and the problem of local minima for most robust estimators has also been taken into account. The resulting network is shown to be highly robust and stable against a high fraction of outliers as well as small perturbations. This demonstrates its superiority in controlling bias and variance of estimators.
['Sheng-Tun Li', 'Ernst L. Leiss']
Towards positive-breakdown radial basis function networks
232,168
Transmission collision is one of the main reasons of performance degradation in dense wireless sensor networks (WSNs). Transmission collision can cause throughput reduction, excessive delay and packet loss. One of the methods to minimize the probability of packet collision is the reduction of the collision area (area around sending nodes where collision may take place). In this study we investigate the problem of collision probability minimization through the use of cooperative transmissions and optimal power allocation in WSNs. We formulate the problem as a constrained optimization problem subject to an outage probability constraint. We determine the optimal transmission power of the source and the relay nodes which minimizes the collision area. Results show that the proposed technique significantly reduces the collision area while keeping the outage performance below the targeted value. Results also show that the proposed technique outperforms the direct transmission system as well as the cooperative system with equal transmission power.
['Fatemeh Mansourkiaie', 'Mohamed Hossam Ahmed', 'Yasser Gadallah']
Minimizing the probability of collision in wireless sensor networks using cooperative diversity and optimal power allocation
499,543
Boosting Higher-Order Correlation Attacks by Dimensionality Reduction.
['Nicolas Bruneau', 'Jean-Luc Danger', 'Sylvain Guilley', 'Annelie Heuser', 'Yannick Teglia']
Boosting Higher-Order Correlation Attacks by Dimensionality Reduction.
846,320
A novel model using support vector regression (SVR) combined with particle swarm optimization (PSO) integrating leave-one-out cross validation (LOOCV) was employed to construct mathematical model for prediction of the magnetic remanence of the NdFeB magnets. The leave-one-out cross validation of SVR model test results show that the mean absolute error doesnot exceed 0.0036, the mean absolute percentage error is only 0.53%, and the correlation coefficient (R 2 ) is as high as 0.839. This investigation suggests that the SVR-LOOCV is not only an effective and practical method to simulate the remanence of NdFeB, but also a powerful tool to optimize designing or controlling the experimental process.
['Wen-De Cheng']
Prediction of magnetic remanence of NdFeB magnets by using novel machine learning intelligence approach — Support vector regression
43,899
We consider the line spectral estimation problem which aims to recover a mixture of complex sinusoids from a small number of randomly observed time domain samples. Compressed sensing methods formulates line spectral estima- tion as a sparse signal recovery problem by discretizing the continuous frequency parameter space into a finite set of gri d points. Discretization, however, inevitably incurs errors and leads to deteriorated estimation performance. In this paper, we propose a new method which leverages recent advances in tensor decomposition. Specifically, we organize the observ ed data into a structured tensor and cast line spectral estimation as a CANDECOMP/PARAFAC (CP) decomposition problem with missing entries. The uniqueness of the CP decomposition allows the frequency components to be super-resolved with infinite precision. Simulation results show that the proposed method provides a competitive estimate accuracy compared with existing state-of-the-art algorithms.
['Jun Fang', 'Linxiao Yang', 'Hongbin Li']
Spectral Compressed Sensing via CANDECOMP/PARAFAC Decomposition of Incomplete Tensors
654,719
Abstract To examine current status and quality of CVD related apps available for download in China, a total of 151 apps from the top popular six app stores were analyzed. Data analysis uncovered a range of issues including missing of key variables in the pre-formatted daily records, no platform for interaction with relevant healthcare professionals and undesirable user-interface design. More importantly, these apps had low levels of adherence to internationally recognized guidelines in CVD management. Overall quality score of these apps was below the average (8.08/20). This study identified areas for improvement concerning the existing CVD related apps. Information may guide the further advancement of CVD related apps and benefit CVD management in China.
['Qian Xiao', 'Yan-ling Wang', 'Liu Sun', 'Sai Lu', 'Ying Wu']
Current Status and Quality Assessment of Cardiovascular Diseases Related Smartphone Apps in China.
832,196
The paper considers the performance analysis of two protocols, carrier sensing multiple access (CSMA) and receiver initiated busy tone multiple access (RI-BTMA) in a distributed wireless ad-hoc network. Carrier sensing and busy tone schemes are used to reduce interference and resolve collisions between randomly distributed nodes. The packet arrival and transmission process is modeled as a regenerative Markov random process. The probability of correct packet detection depends on the SINR of the received packet. The SINR of the packet, the signal received during channel sensing by transmitter nodes in the CSMA protocol, and the busy tone signal magnitude at transmitter nodes using the RI-BTMA protocol are modeled as a random variable whose cumulative distribution function is derived. The throughput of the two systems for fixed outage probability is analytically derived. Analytical results are verified using discrete time step simulation.
['Vishal Parikh', 'Didem Kivanc-Tureli', 'Uf Tureli']
Performance analysis of CSMA and RI-BTMA in an ad hoc network
125,089
Copyright (c) 1997 Elsevier Science B.V. All rights reserved. A language L over the Cartesian product of component alphabets is called projective if it is closed under projections, i.e., together with each word α∈L, it contains all the words that have the same projections up to stuttering as α. We prove that the projective languages are precisely the languages obtained using parallel composition and intersection from stuttering-closed component languages in each of the following classes of languages: regular, star-free regular, ω-regular and star-free ω-regular. Languages of these classes can also be seen as properties of various temporal logics which are used to specify properties of concurrent systems. In particular, the star-free ω-regular languages coincide with properties expressed using Propositional Linear Temporal Logic. Some uses of projective properties for specification and verification of programs are studied.
['Doron A. Peled']
On projective and separable properties
274,306
A new model, called Local-DNN, is proposed for the gender recognition problem.The model is based on local features and deep neural networks.The local contributions are combined in a voting scheme for the final classification.The model obtains state-of-the-art results in two wild face image datasets. Display Omitted Deep learning methods are able to automatically discover better representations of the data to improve the performance of the classifiers. However, in computer vision tasks, such as the gender recognition problem, sometimes it is difficult to directly learn from the entire image. In this work we propose a new model called Local Deep Neural Network (Local-DNN), which is based on two key concepts: local features and deep architectures. The model learns from small overlapping regions in the visual field using discriminative feed-forward networks with several layers. We evaluate our approach on two well-known gender benchmarks, showing that our Local-DNN outperforms other deep learning methods also evaluated and obtains state-of-the-art results in both benchmarks.
['Jordi Mansanet', 'Alberto Albiol', 'Roberto Paredes']
Local Deep Neural Networks for gender recognition
562,505
The relative motion between the camera and the scene leads to blurring of images. By appealing to traveling wave equations, the restoration of the motion-blurred images can be achieved and a criterion of objective quality assessment can be obtained for these restored versions even without any a priori knowledge of the original image. Most significantly, this criterion actually applies to any categories of image restoration methods as long as the image blur is caused by a linear uniform motion with good sensitivity and consistency.
['Li-Dong Cai']
Objective assessment to restoration of global motion-blurred images using traveling wave equations
341,742
It is already clarified throughout studies of passive dynamic walking mechanisms that the common nec essary condition for dynamic gait generation comes from the requirement on mechanical energy restoration. Until now we have treated only rotational joints of the robot, whereas in this paper we consider a novel dynamic gait generation method based on mechanical energy restoration by parametric excitation using telescopic leg actuation. We first introduce a simple walking model and a control law for the telescopic leg motion, and show the typical walking pattern by numerical simulations. We then analyze the gait performance by adjusting some control and physical parameters. In addition, some extensions of the mechanism and control applications are investigated.
['Fumihiko Asano', 'Zhiwei Luo', 'Sang-Ho Hyon']
Parametric Excitation Mechanisms for Dynamic Bipedal Walking
379,574
A new error correction scheme based on a brain-inspired learning algorithm, called Recurrent Neural Network (RNN), is proposed for resilient and efficient intra-chip data transmission. RNN has a feature to find partially-clustered time-series data stream and predict the next input data from previous input data stream. By utilizing this feature, a novel top-down error correction approach which considers the “context” included in the data stream and predicts original data by an acquired knowledge can be realized. In this paper, the performance of a RNN/BCH-hybrid error correction scheme for reducing the effect of false-positive detection is demonstrated through an experimental evaluation using a general purpose microprocessor.
['Masanori Natsui', 'Naoto Sugaya', 'Takahiro Hanyu']
A study of a top-down error correction technique using Recurrent-Neural-Network-based learning
931,934
A Multiple Trust Paths Selection Tool in Contextual Online Social Networks
['Linlin Ma', 'Guanfeng Liu', 'Guohao Sun', 'Lei Li', 'Zhixu Li', 'An Liu', 'Lei Zhao']
A Multiple Trust Paths Selection Tool in Contextual Online Social Networks
759,438
An important problem when modeling gene networks lies in the identification of parameters, even if we consider a purely discrete framework as the one of Ren\'e Thomas. Here we are interested in the exhaustive search of all parameter values that are consistent with observed behaviors of the gene network. We present in this article a new approach based on Hoare Logic and on a weakest precondition calculus to generate constraints on possible parameter values. Observed behaviors play the role of "programs" for the classical Hoare logic, and computed weakest preconditions represent the sets of all compatible parameterizations expressed as constraints on parameters. Finally we give a proof of correctness of our Hoare logic for gene networks as well as a proof of completeness based on the computation of the weakest precondition.
['Gilles Bernot', 'Jean-Paul Comet', 'Zohra Khalis', 'Adrien Richard', 'Roux O']
A Genetically Modified Hoare Logic
574,000
This paper describes the development of an effective and efficient Hierarchical and Parallel Branch-and-Bound Ensemble Selection (H&PB&BEnS) algorithm. Using the proposed H&PB&BEnS, ensemble selection is accomplished in a divisional, parallel, and hierarchical way. H&PB&BEnS uses the superior performance of the Branch-and-Bound (B&B) algorithm in relation to small-scale combinational optimization problems, whilst also managing to avoid “the curse of dimensionality” that can result from the direct application of B&B to ensemble selection problems. The B&B algorithm is used to select each partitioned subensemble, which enhances the predictive accuracy of each pruned subsolution, and then the working mechanism of H&PB&BEnS improves the diversity of the ensemble selection results. H&PB&BEnS realizes layer-wise refinement of the selected ensemble solutions, which enables the classification performance of the selected ensembles to be improved in a layer-by-layer manner. Empirical investigations are conducted using five benchmark classification datasets, and the results verify the effectiveness and efficiency of the proposed H&PB&BEnS algorithm.
['Qun Dai', 'ChangSheng Yao']
A hierarchical and parallel branch-and-bound ensemble selection algorithm
846,191
User Generated Content Oriented Chinese Taxonomy Construction
['Jinyang Li', 'Chengyu Wang', 'Xiaofeng He', 'Rong Zhang', 'Ming Gao']
User Generated Content Oriented Chinese Taxonomy Construction
756,082
Robotics: Probabilistic Methods for State Estimation and Control.
['Wolfram Burgard', 'Cyrill Stachniss']
Robotics: Probabilistic Methods for State Estimation and Control.
791,623
In biological applications and systems where even the smallest details have a meaning, CCD cameras are mostly preferred and they hold most of the market share despite their high costs. In this paper, we propose a custom-designed CMOS camera to compete with the default CCD camera of an inverted microscope for fluorescence imaging. The custom-designed camera includes a commercially available mid-performance CMOS image sensor and a Field-Programmable Gate Array (FPGA) based hardware platform (FPGA4U). The high cost CCD camera of the microscope is replaced by the custom-designed CMOS camera and the two are quantitatively compared for a specific application where an Estrogen Reception (ER) expression in breast cancer diagnostic samples that emits light at 665nm has been imaged by both cameras. The gray-scale images collected by both cameras show a very similar intensity distribution. In addition, normalized white pixels after thresholding resulted in 4.96% for CCD and 3.38% for CMOS. The results and images after thresholding show that depending on the application even a mid-performance CMOS camera can provide enough image quality when the target is localization of fluorescent stained biological details. Therefore the cost of the cameras can be drastically reduced while benefiting from the inherent advantages of CMOS devices plus adding more features and flexibility to the camera systems with FPGAs.
['Gözen Köklü', 'Julien Michel Ghaye', 'René Beuchat', 'Giovanni De Micheli', 'Yusuf Leblebici', 'Sandro Carrara']
Quantitative comparison of commercial CCD and custom-designed CMOS camera for biological applications
427,238
Systems featuring a load-store interface to persistent memory (PM) are expected soon, making in-memory persistent data structures feasible. Ensuring persistent data structure recoverability requires constraints on the order PM writes become persistent. But, current memory systems reorder writes, providing no such guarantees. To complement their upcoming 3D XPoint memory, Intel has announced new instructions to enable programmer control of data persistence. We describe the semantics implied by these instructions, an ordering model we call synchronous ordering. Synchronous ordering (SO) enforces order by stalling execution when PM write ordering is required, exposing PM write latency on the execution critical path. It incurs an average slowdown of 7.21x over volatile execution without ordering in PM-write-intensive benchmarks. SO tightly couples enforcing order and flushing writes to PM, but this tight coupling is unneeded in many recoverable software systems. Instead, we propose delegated ordering, wherein ordering requirements are communicated explicitly to the PM controller, fully decoupling PM write ordering from volatile execution and cache management. We demonstrate that delegated ordering can bring performance within 1.93x of volatile execution, improving over SO by 3.73x.
['Aasheesh Kolli', 'Jeff Rosen', 'Stephan Diestelhorst', 'Ali G. Saidi', 'Steven Pelley', 'Sihang Liu', 'Peter M. Chen', 'Thomas F. Wenisch']
Delegated persist ordering
965,639
An Attribute-Based Protection Model for JSON Documents
['Prosunjit Biswas', 'Ravi S. Sandhu', 'Ram Krishnan']
An Attribute-Based Protection Model for JSON Documents
891,007
Questioning strangers about critical medical decisions: “What happens if you have sex between the HPV shots?”
['Lynn Westbrook', 'Yan Zhang']
Questioning strangers about critical medical decisions: “What happens if you have sex between the HPV shots?”
732,949
In this paper we describe an original method for the 3D free form object tracking in monocular vision. The main contribution of this article is the use of the skeleton of an object in order to recognize, locate and track this object in real time. Indeed, the use of this kind of representation made it possible to avoid difficulties related to the absence of prominent elements in free form objects (which makes the matching process easier). The skeleton is a lower dimension representation of the object, it is homotopic and it has a graph structure. This allowed us to use powerful tools of the graph theory in order to perform matching between scene objects and models (recognition step). Thereafter, we used skeleton extremities as interest points for the tracking. Keywords: Tracking, 3D free form object, Skeletonization, Graph matching.
['Djamel Merad', 'Jean-Yves Didier', 'Mihaela Scuturici']
Tracking 3D free form object in video sequence
463,748