Entity

Time filter

Source Type


Gould S.,National School in Computer Science
IEEE Transactions on Pattern Analysis and Machine Intelligence | Year: 2015

Markov random fields containing higher-order terms are becoming increasingly popular due to their ability to capture complicated relationships as soft constraints involving many output random variables. In computer vision an important class of constraints encode a preference for label consistency over large sets of pixels and can be modeled using higher-order terms known as lower linear envelope potentials. In this paper we develop an algorithm for learning the parameters of binary Markov random fields with weighted lower linear envelope potentials. We first show how to perform exact energy minimization on these models in time polynomial in the number of variables and number of linear envelope functions. Then, with tractable inference in hand, we show how the parameters of the lower linear envelope potentials can be estimated from labeled training data within a max-margin learning framework. We explore three variants of the lower linear envelope parameterization and demonstrate results on both synthetic and real-world problems. © 2015 IEEE. Source


Movahedi Z.,University Pierre and Marie Curie | Ayari M.,University Pierre and Marie Curie | Ayari M.,National School in Computer Science | Langar R.,University Pierre and Marie Curie | Pujolle G.,University Pierre and Marie Curie
IEEE Communications Surveys and Tutorials | Year: 2012

Autonomic network management is a promising approach to reduce the cost and the complexity of managing network infrastructures. It attempts to lead the human administrator out of the network control loop, leaving the management tasks to be performed by the network itself. Due to its important implication on automating management systems, this area has attracted a growing attention from both academia and industry. In this paper, we provide a holistic view of autonomic architecture proposals and the evaluation metrics existing so far. Based on this, we identify some new criteria important to the autonomic architectures. Finally, we compare different existing autonomic architectures and describe the pros and cons of each one regarding to the network management and performances. © 1998-2012 IEEE. Source


Kieu T.D.,National School in Computer Science | Chang C.-C.,Feng Chia University
Expert Systems with Applications | Year: 2011

Recently, Zhang and Wang proposed a steganographic scheme by exploiting modification direction (EMD) to embed one secret digit d in the base-(2 × n + 1) notational system into a group of n cover pixels at a time. Therefore, the hiding capacity of the EMD method is log2(2 × n + 1)/n bit per pixel (bpp). In addition, its visual quality is not optimal. To overcome the drawbacks of the EMD method, we propose a novel steganographic scheme by exploiting eight modification directions to hide several secret bits into a cover pixel pair at a time. By this way, the proposed method can achieve various hiding capacities of 1, 2, 3, 4, and 4.5 bpp and good visual qualities of 52.39, 46.75, 40.83, 34.83, and 31.70 dB, respectively. The experimental results show that the proposed method outperforms three recently published works, namely Mielikainen's, Zhang and Wang's, and Yang et al.'s methods. © 2011 Published by Elsevier Ltd. Source


El Korbi I.,National School in Computer Science | Zeadally S.,University of Kentucky
Ad Hoc Networks | Year: 2014

We propose a sensor node relocation approach in wireless sensor networks to maintain connectivity between a Region Of Interest (ROI) where the sensor nodes are initially deployed and a Center Of Interest (COI) outside the ROI where a particular event happens. Our proposed approach, called Chain Based Relocation Approach (CBRA), aims to relocate a minimum number of redundant sensors from their initial positions within the ROI towards the COI to maintain the connectivity between the ROI and the COI. CBRA uses steps which determine the redundant nodes' set, the propagation of the COI coordinates within the ROI and then the selection and the relocation of the redundant nodes towards the COI. The selection of the redundant nodes is based on an average energy consumption model to balance the energy consumption among the sensor nodes when they are relocated depending on their initial and final positions. We evaluate the performance of CBRA using performance metrics such as energy consumption, the number of relocated nodes, relocation time and number of transmitted messages. Sensor nodes are relocated using a chain-based method between the ROI and the COI. In addition, if one relocated sensor node fails, the connectivity between the COI and the ROI is affected. To address this possible failure, we propose a fault tolerant recovery procedure to repair the route between the COI and the ROI. Finally, we compare the performance of CBRA with two other approaches. © 2014 Elsevier B.V. All rights reserved. Source


Chakchouk N.,National School in Computer Science
IEEE Communications Surveys and Tutorials | Year: 2015

The great advances made in the wireless technology have enabled the deployment of wireless communication networks in some of the harshest environments such as volcanoes, hurricane-affected regions, and underground mines. In such challenging environments suffering from the lack of infrastructure, traditional routing is not efficient and sometimes not even feasible. Moreover, the exponential growth of the number of wireless connected devices has created the need for a new routing paradigm that could benefit from the potentials offered by these heterogeneous wireless devices. Hence, in order to overcome the traditional routing limitations, and to increase the capacity of current dynamic heterogeneous wireless networks, the opportunistic routing paradigm has been proposed and developed in recent research works. Motivated by the great interest that has been attributed to this new paradigm within the last decade, we provide a comprehensive survey of the existing literature related to opportunistic routing. We first study the main design building blocks of opportunistic routing. Then, we provide a taxonomy for opportunistic routing proposals, based on their routing objectives as well as the optimization tools and approaches used in the routing design. Hence, five opportunistic routing classes are defined and studied in this paper, namely, geographic opportunistic routing, link-state-aware opportunistic routing, probabilistic opportunistic routing, optimization-based opportunistic routing, and cross-layer opportunistic routing. We also review the main protocols proposed in the literature for each class. Finally, we identify and discuss the main future research directions related to the opportunistic routing design, optimization, and deployment. © 2015 IEEE. Source

Discover hidden collaborations