|
ICTEAM Young Researchers' DayThird edition : Wednesday May 22, 2013
 In 2013, the ICTEAM Young Researchers' Day took place at the Ferme de Profondval in Court-Saint-Etienne, very close to Louvain-la-Neuve.
Programme At Profondval
At Louvain-la-Neuve 19h00 BBQ Abstracts Microsensors for the Health, from Earth to Space by Prof. Laurent Francis (ELEN) : Microsensors have an increasing interest for physical or (bio-)chemical sensing with better performances, lower power requirements, and lower materials consumption with respect to their macroscopic counterparts, when these simply exist. In this presentation, I will focus mainly on the miniaturized sensors that are used for measuring a health status for different situations. In a first situation, in the biomedical field, where the human health can be interrogated with microsensors for detecting pathogens or the breathing activity of patients. In a second situation, in the aeronautics and space field, where the structural health monitoring is taking place and where, this time, microsensors are exposed to harsh environmental conditions. For both situations the presentation will cover the design, fabrication and characterization aspects of the microsensors Miniaturized sensing platform for high temperature environments by Nicolas André (ELEN) : Humidity and UV/blue light generated by the methane combustion are monitored at the source by sensors, electronics and packaging resistant to the harsh conditions met in domestic boilers. Significant energy consumption savings as well as waste production reducing are targeted by this improved knowledge of the combustion conditions. Joint spectral characteristics: a tale of three disciplines by Prof. Raphaël Jungers and Pierre-Yves Chevalier (INMA) : Joint spectral characteristics describe the stationary behavior of a discrete time linear switching system. Well, that's what an electrical engineer would say. A mathematician would say that they characterize the asymptotic behavior of a semigroup of matrices, and a computer scientist would perhaps see them as describing languages generated by automata. Because of their connections with these wide research topics, joint spectral characteristics have been at the center of rich and diverse research efforts in recent years. They are notoriously very hard to compute (NP-hardness, Undecidability, etc. are the rule rather than the exception), but it turns out that one can often get around these difficulties, and modern optimization techniques seem particularly useful for studying them. We will survey and connect several powerful and interesting results. We will present applications, ranging from wireless control protocols to viral diseases treatment, malicious agents tracking, and consensus of multi-agent systems. Solving Vehicle Routing Problems by Prof. Yves Deville and Florence Massen (INGI) : Vehicle Routing Problems (VRPs) cover a large set of problems and have many industrial and commercial applications. VRPs have received a great deal of attention since as early as the 1960's. While initially only basic variants have been considered, during the following decades research focused on more complex variants, such as problems with time windows or with pick-up and delivery. In the first part of this talk, we will present a basic VRP and several problem variants, as well as different approaches for solving these problems. Research in recent years has tackled so-called Rich Vehicle Routing Problems. These problems strive to give a more realistic representation of problems encountered in the real-world. Rich problems often require that the combination of different complicating constraints be handled. In a second part of the talk, we will introduce a generalized VRP variant, the VRP with Black Box Feasibility (VRPBB). This generalization allows to accommodate emerging routing problems with hard side constraints to be respected by every route for which no efficient exact feasibility check is available. Typical applications would be combinations of routing with loading or routing with scheduling. We will show how a heuristic column generation approach can be used for solving the VRPBB. Green Systems-on-a-Chip for a Sustainable Internet-of-Things by Prof. David Bol (ELEN) : The vision of the Internet-of-Things (IoT) calls for the deployment of trillions of wireless sensor nodes (WSNs) in our environment with new exciting applications in fields such as energy savings, habitat, transportation, industrial production, healthcare, entertainment and security. A sustainable deployment of such a large number of electronic systems needs to be addressed with a Design-for-the-Environment (DfE) approach. In the Electronics Circuits and Systems (ECS) group, we aim at designing of ultra-low-power and ultra-compact Systems-on-a-Chip i.e. green SoCs to minimize the carbon footprint of WSN production for a massive yet sustainable deployment of the IoT. Connecting people: communities in mobile phone networks by Prof. Vincent Blondel (INMA) : I will overview several recent results on network analysis with a special emphasis on community detection and on the analysis of mobile phone datasets. In particular, I will describe a simple and efficient method developed within ICTEAM - now known as the "Louvain method" - for the detection of communities. The Louvain method - used by LinkedIn for its visualization application InMaps - can be routinely used for analyzing networks with billions of nodes or links. I will analyze the communities obtained on a nationwide dataset of criminal records, as well as on a social network constructed from mobile phone communications in Belgium and in France on periods covering several months. I will also describe applications of mobile phone dataset analysis for a range of applications such as urban planning, traffic optimization, monitoring of development policy, crisis management, and control of epidemics. With these applications in mind I will demonstrate results obtained in the "Data for Development" (D4D) challenge organized jointly with Orange on the analysis of mobile phone datasets from an African country and for development purposes. This is joint work with a number of past and present students and collaborators; including Gautier Krings, Yves-Alexandre de Montjoye, Adeline Decuyper and Pierre Deville. Can You See It, Can You Do It: Model-Checking of Observation and Control by Prof. Charles Pecheur (INGI) : Have you ever been confused as to when and why your new car decides to lock the doors, keep lights on or switch off the radio? Can you transpose that situation with the auto-pilot of an Airbus? In the Louvain Verification Lab (LVL) team, we are using model-checking to analyse observation and control properties of computer systems, under different angles. Model-checking is an analysis technique based on the systematic exploration of the state space of a dynamic system. We use it to check that available observations are sufficient to properly detect and diagnose failures, or to synthesize abstract models suitable for training human operators. More generally, knowledge and strategies of different agents within a system can be expressed and analyzed with suitable logics and model-checking algorithms. In support of these activities, we are also developing a Python-based programmable verification platform and visualization tools to display and interactively explore the rich results of these analyses. This is joint work with LVL researchers Simon Busard and Sébastien Combéfis. Discover the power of the Body Language by Vincent Geeraerd (Le Soleil Consulting) : You often present subjects or projects, you have to catch the attention of participants. Words seem to be enough.
Do you know that most of meaning in human communication is given by our body language: our smile, our face, eye contact, movements, … are speaking for us. They give very much more information than words.
Through a short and effective presentation, we will show you how to discover and improve the power of your body language.
NEGF Computational study of Advanced Nanoscale FET Biosensors for Single Nanoscale Analyte Detection - Aryan Afzalian (ELEN) Advanced nanoscale devices like Si nanowires or graphene nanoribbons transistors are very promising as transducers in particular for label-free sensing and detection of biological species. We have used a dedicated NEGF microscopic quantum simulation platform to investigate the ultimate limit of detection of such advanced nanoscale FETs. The question we sought to answer was what are the true opportunities and challenges to reach single nanoscale analyte detection (e.g. a small DNA chain) where the amount of charges to detect is small and its impact on the transducer behavior can be further reduced due to ion and oxide charges screening.
UWB Imaging Radar System Including 3D Circular Vivaldi Antennas Array - Khaldoun Alkhlifeh (ELEN)
Designing and fabrication a 3D single Vivaldi antenna will be presented. The measurement results of fabrication a 3D circular Vivaldi antennas array will be exposed.
Incremental classification for image segmentation - Guillaume Bernard (ELEN)
Image segmentation problems can be solved with classification algorithms. Their use is limited to features derived from intensities of pixels or patches. Features such as contiguity of two regions cannot be considered without prior knowledge of one of the two class labels. We propose an incremental classifier that works in a space where features are progressively evaluated.
Train&Align: a new online tool for automatic phonetic alignment - Sandrine Brognaux (INGI)
Several automatic phonetic alignment tools have been proposed in the literature. They usually rely on pre-trained speaker-independent models to align new corpora. Their drawback is that they cover a very limited number of languages and might not perform properly for different speaking styles. This paper presents a new tool for automatic phonetic alignment available online. Its specificity is that it trains the model directly on the corpus to align which makes it applicable to any language and speaking style. Experiments on three corpora show that it provides results comparable to other existing tools. It also allows the tuning of some training parameters. The use of tied-state triphones for example shows further improvement of about 1.5% for a 20 ms threshold. A manually-aligned part of the corpus can also be used as bootstrap to improve the model quality. Alignment rates were found to significantly increase up to 20% using only 30 seconds of bootstrapping data.
Securely Solving Simple Combinatorial Graph Problems - Edouard Cuvelier (ELEN)
We investigate the problem of solving traditional combinatorial graph problems using secure multi-party computation techniques, focusing on the shortest path and the maximum flow problems. To the best of our knowledge, this is the first time these problems have been addressed in a general multi-party computation setting. Our study highlights several complexity gaps and suggests the exploration of various trade-offs, while also offering protocols that are efficient enough to solve real-world problems.
Grip force adaptation to dynamic torques during object manipulation in weightlessness - Thibault Giard (INMA)
In many situations we manipulate objects grasped away from their center of mass which induces torques applied tangentially to the grip surfaces when the object is accelerated. When the object configuration induces a tangential torque relative to the grip axis we previously showed that the grip force exhibits a gradual and slow decrease as subjects learned to manipulate the unbalanced object (Crevecoeur et al. 2011). The specific contribution of a tangential torque in the sensorimotor control of grip force remains unexplored and very hard to dissociate from usual loads experienced during object manipulation.
Representations and Invariants of Open Reacting Systems - Nicolas Hudon (INMA)
This contribution considers the problem of representing and characterizing open reacting systems in the thermodynamical phase space by taking into account reaction invariants. The objective is to get a more precise picture of the system structure.
Experimental Evaluation of Multi-user Separation in Urban Microcellular Networks - Nizabat Khan (ELEN)
Multi-user MIMO channels have become more popular due to their inherent potential for capacity improvement, but at the cost of increasing interference. The latter is reduced when the channel matrices show a sufficient spatial separation. In this work, MU- MIMO channels separation is characterized using experimental measurements in an urban microcellular scenario. A base station (equipped with a polarized directional antenna) transmits to a receiver comprising of a vertical antenna array of 8 elements. To characterize the multi-user separation, three metrics are evaluated namely: (i) shadow fading correlation (SFC), (ii) correlation matrix distance (CMD) and (iii) spectral divergence (SF). The experimental evaluation shows that different users placed at 5-10 m distance can have acceptable separation in microcellular networks.
Developments in the theory of randomized shortest paths - Ilkka Kivimäki (INGI)
There have lately been several suggestions for parametrized distances between nodes of a graph that generalize the shortest path distance and the commute time or resistance distance. This poster presents developments in the theory of one such family of distances known as the randomized shortest path (RSP) dissimilarity. We derive a new definition of a distance measure inspired by the RSP dissimilarity that we call the free energy distance. We also make a comparison between the RSP dissimilarity the free energy distance and other parametrized graph node distances.
Providing Energy-neutral Ancillary Services from Responsive Loads - Arnaud Latiers (INMA)
A suggestion for balancing the electricity grid in real-time through the use of flexibility in industrial electricity consumption.
Semi-Supervised Classification Through the Bag-of-Paths Group Betweenness - Bertrand Lebichot (INGI)
The goal is to tackle semisupervised classification problems on weighted directed graphs. The objective of semi-supervised classification is to assign a label to unlabeled nodes using the whole topology of the graph and the labeled nodes at our disposal. A node receives a high group betweenness if it has a large probability of appearing on paths connecting two nodes of the same classe. Unlabeled nodes are then classified according to the class showing the highest group betweenness. Experiments on various real-world data sets show that BoP group betweenness outperforms all the tested state-of-the-art methods. The benefit of the BoP betweenness is particularly noticeable when only a few labeled nodes are available.
Fabrication of a Nanowire Field Effect Transistor for Biosensor Application - Achim Müller (ELEN)
This poster shows a part of my master thesis. The goal of the thesis is to fabricate NW-FET devices with very low dimensions. In this case that means a height and a width down to 5 nm and a length down to 50 nm. The fabrication will be realized by a combination of optical lithography and e-beam lithography. The etching of the NW is realized by a NH4OH solution. The reason for such a solution is to reach low surface roughness
The iMagX research project - Jonathan Orban de Xivry (ELEN)
iMagX is a joint project between the UCL (ICTEAM/ELEN and IREC/MIRO) and IBA world leader in proton therapy. Our interdisciplinary team of 20 engineers computer scientists physicists and PhDs works at the intersection of the scientific clinical and industrial worlds in order to create innovative imaging solutions to improve cancer treatment in proton therapy and radiotherapy.
Passive RFID systems: eigen-mode analysis - Donia Oueslati (ELEN)
A numerical analysis method based on a combination of TE and TM eigen-modes is proposed for the study of the current distribution in a microstrip patch antenna devoted to passive Radio Frequency Identification (RFID) systems. The analysis is based on eigen-value and eigen-vector decomposition of the Method-of-Moments impedance matrix. The process of projection of incident plane waves on those eigen-modes is studied and a methodology is delineated to selectively excite certain modes using multiple plane waves.
Synthesis of bio-inspired multilayer photonic crystal using ALD - Olivier Poncelet (ELEN)
It is a poster on my own work. It will explain in detail a process for the fabrication of a bio-inspired photonic crystal observed on the Cicendela chinensis cuticule. This kind of structures are relevant to fight again the counterfeiting because of their polarisation of light capacities.
Organizers AsCII (Association des chercheurs de l'institut ICTEAM) : Adeline Decuyper, Sébastien Combéfis.
|
27/05/2013
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
