• Publicationsclose
• 2013-2022close
• Conference object close
• CA close
• AT close
• English close
• Mémoires en Sciences de l'Information et de la Communication close

International audience; A deep learning surrogate for the direct numerical prediction of two-dimensional acoustic waves propagation and scattering with obstacles is developed through an auto-regressive spatio- temporal convolutional neural network. A single database of high-fidelity lattice Boltzmann temporal simulations is employed in the training of the network, achieving accurate predictions for long simulation times for a variety of test cases, representative of bounded and unbounded configurations. The capacity of the network to extrapolate outside the manifold of examples seen during the training phase is demonstrated by the obtaining of accurate acoustic predic- tions for relevant applications, such as the scattering of acoustic waves on an airfoil trailing edge, an engine nacelle or in-duct propagation. The method is tested for two types of input normalizations, coupled with an a-posteriori correction which improves the acoustic energy conservation of the predictions. The use of an adaptive local normalization along with the physics-based energy conservation results in an error reduction for all the studied cases.

Renal compartment segmentation on CT images targets on extracting the 3D structure of renal compartments from abdominal CTA images and is of great significance to the diagnosis and treatment for kidney diseases. However, due to the unclear compartment boundary, thin compartment structure and large anatomy variation of 3D kidney CT images, deep-learning based renal compartment segmentation is a challenging task. We propose a novel weakly supervised learning framework, Cycle Prototype Network, for 3D renal compartment segmentation. It has three innovations: (1) A Cycle Prototype Learning (CPL) is proposed to learn consistency for generalization. It learns from pseudo labels through the forward process and learns consistency regularization through the reverse process. The two processes make the model robust to noise and label-efficient. (2) We propose a Bayes Weakly Supervised Module (BWSM) based on cross-period prior knowledge. It learns prior knowledge from cross-period unlabeled data and perform error correction automatically, thus generates accurate pseudo labels. (3) We present a Fine Decoding Feature Extractor (FDFE) for fine-grained feature extraction. It combines global morphology information and local detail information to obtain feature maps with sharp detail, so the model will achieve fine segmentation on thin structures. Our extensive experiments demonstrated our great performance. Our model achieves Dice of \(79.1\%\) and \(78.7\%\) with only four labeled images, achieving a significant improvement by about \(20\%\) than typical prototype model PANet [16].

International audience; Les communications sans contact sont omniprésentes dans notre quotidien, allant des badges de contrôle d'accès au passeport électronique. Ces systèmes sont sensibles aux attaques par relais, dans lesquelles un adversaire transfère simplement les messages entre le prouveur et le vérifieur pour usurper l'identité du prouveur. Les protocoles délimiteurs de distance (distance-bounding) ont été ont ntroduits pour contrer ces attaques en assurant une borne sur la distance entre le prouveur et le vérifieur grâcè a la mesure du temps des communications. Par la suite de nombreux travaux ont amélioré la sécurité de ces protocoles, mais ont aussi cherché à assurer le respect de la vie privée face à des adversaires actifs et également face à des vérifieurs malicieux. En particulier, une menace difficile à prévenir est la fraude terroriste, où un prouveur lointain coopère avec un complice proche pour tromper le vérifieur. La contre-mesure usuelle pour cette menace est de rendre impossible l'action du complice sans l'aide du prouveur lointain, à moins que le prouveur ne lui donne suffisamment d'information pour qu'il retrouve sa clef privée et puisse ainsi toujours se faire passer pour le prouveur. Dans cet article, nous proposons une nouvelle approche où le prouveur ne révèle pas sa clef privée mais utilise une clef de session avec une signature de groupe, la rendant ainsi utilisable plusieurs fois. Ceci permet à un adversaire d'usurper l'identité du prouveur sans même connaître sa clef de signature. Grâce à cette approche nous proposons SPADE le premier protocole de délimiteur de distance qui est anonyme, révocable et formellement prouvé sûr. Mots-clefs : Protocole délimiteur de distance (Distance Bounding), Sécurité, résitance à la fraude terroriste.

In this paper, a scheduling protocol for electric vehicle (EV) home charging with time of use pricing is introduced. This work addresses the problem of EVs charging at home by adopting an appropriate charging process protocol over Power Line Communications (PLC). The scheduling protocol is aimed at minimizing peak loads on distribution feeders due to multiple EVs charging while using a time-of-use pricing policy. Energy efficiency and performance are both taken into account. An appropriate analytical formulation of the scheduling problem is given together with the proposed scheduling protocol. Simulations demonstrate the effectiveness of the proposed approach in minimizing peak loads while satisfying the defined constraints.

Part 7: Digital Governance; International audience; Infrastructural development is a significant determinant of economic growth. It remains an elusive pursuit for many developing economies suffering from public infrastructural project failures. Although the causes of these failures are identifiable, they remain persistent. Government corruption has been identified as the primary cause of project failures amidst a host of other causal factors, spurred by the ambiguity in public service administration. These factors heighten capital expenditures and hence, the need for more transparent systems in public infrastructural project planning and -delivery. This research uses a case-study methodology to examine the importance of public involvement in addressing the causes of failures in public infrastructural project planning and -delivery. Using Nigeria as a case, the findings from conducted interviews and a document review support the proposition of a technologically collaborative approach in addressing the causes of public infrastructural project failures. The institutionalization of transparency-enhancing blockchain systems are vital in government and public involvement in the processes of public infrastructural project planning and -delivery.

Part 8: Pointing, Touch, Gesture and Speech-Based Interaction Techniques; International audience; Eyewear displays allow users to interact with virtual content displayed over real-world vision, in active situations like standing and walking. Pointing techniques for eyewear displays have been proposed, but their social acceptability, efficiency, and situation awareness remain to be assessed. Using a novel street-walking simulator, we conducted an empirical study of target acquisition while standing and walking under different levels of street crowdedness. We evaluated three phone-based eyewear pointing techniques: indirect touch on a touchscreen, and two in-air techniques using relative device rotations around forward and a downward axes. Direct touch on a phone, without eyewear, was used as a control condition. Results showed that indirect touch was the most efficient and socially acceptable technique, and that in-air pointing was inefficient when walking. Interestingly, the eyewear displays did not improve situation awareness compared to the control condition. We discuss implications for eyewear interaction design.

International audience; Sliced Inverse Regression (SIR) has been extensively used to reduce the dimension of the predictor space before performing regression. SIR is originally a model free method but it has been shown to actually correspond to the maximum likelihood of an inverse regression model with Gaussian errors. This intrinsic Gaussianity of standard SIR may explain its high sensitivity to outliers as observed in a number of studies. To improve robustness, the inverse regression formulation of SIR is therefore extended to non-Gaussian errors with heavy-tailed distributions. Considering Student distributed errors it is shown that the inverse regression remains tractable via an Expectation- Maximization (EM) algorithm. The algorithm is outlined and tested in the presence of outliers, both in simulated and real data, showing improved results in comparison to a number of other existing approaches.

International audience; The development of virtualization techniques enables an architectural shift in mobile networks, where resource allocation, or even signal processing, become software functions hosted in a data center. The centralization of computing resources and the dynamic mapping between baseband processing units (BBUs) and remote antennas (RRHs) provide an increased flexibility to mobile operators, with important reductions of operational costs. Most research efforts on Cloud Radio Access Networks (CRAN) consider indeed an operator perspective and network-side performance indicators. The impact of such new paradigms on user experience has been instead overlooked. In this paper, we shift the viewpoint, and show that the dynamic assignment of computing resources enabled by CRAN generates a new class of mobile terminal handover that can impair user quality of service. We then propose an algorithm that mitigates the problem, by optimizing the mapping between BBUs and RRHs on a time-varying graph representation of the system. Furthermore, we show that a practical online BBU-RRH mapping algorithm achieves results similar to an oracle-based scheme with perfect knowledge of future traffic demand. We test our algorithms with two large-scale real-world datasets, where the total number of handovers, compared with the current architectures, is reduced by more than 20%. Moreover, if a small tolerance to dropped calls is allowed, 30% less handovers can be obtained.

Integrating wireless sensor networks (WSNs) in power substations for the future smart grid is growing in interest. Nevertheless, high voltage (HV) substations are harsh environments. In particular, impulsive noise needs to be taken into account. To tackle the constraints of these environments, we propose in this paper an efficient wideband channel coding scheme. The proposed approach consists in a robust physical layer based on the integration of several very interesting error correcting codes with Orthogonal Frequency Division Multiplexing (OFDM). Using real measurements of impulsive noise, the impact of rank metric (RC), Low Rank Parity Check codes (LRPC) and polar codes is evaluated in terms of BER and PER in a realistic multipath channel. The results show that using this coding scheme is very efficient in mitigating the bursty nature of impulsive noise while having a quite low level of complexity.

International audience