Publications

2017

• Speech Intelligibility Improvement in Car Noise Environment by Voice Transformation
  
  • Nathwani Karan
  • Richard Gael
  • Bertrand David
  • Prablanc Pierre
  • Roussarie Vincent
  Speech Communication, Elsevier : North-Holland, 2017 (91), pp.17-27. The typical application targeted by this work is the intelligibility improvement of speech messages when rendered in car noise environment (radio, message alerts,...). The main idea of this work is to transform the original speech to “Lombard” speech or more precisely to simulate some of the strategies followed by humans to render their speech clearer when they are surrounded by noise. Three main effects are considered in this work, namely non uniform-time scale modification, formant shifting and a combination of these modifications along with energy redistribution between speech regions. All effects are studied with specific transformations for voiced and unvoiced segments. The proposed modifications are then evaluated by means of subjective and objective tests. The results of these tests conducted with normal hearing and impaired listeners demonstrate the potential of the selected transformations for voice intelligibility improvement
• Stochastic Rank-1 Bandits
  
  • Wen Zheng
  • Kataryia Sumeet
  • Kveton Bratislav
  • Szepesvari Csaba
  • Vernade Claire
  , 2017.
• A quantitative analysis of digital library user behaviours based on access logs
  
  • Nouvellet Adrien
  • d'Alché-Buc Florence
  • Beaudouin Valérie
  • Prieur Christophe
  • Roueff François
  , 2017.
• Collapsing the layers: 6Stor, a scalable and IPv6-centric distributed storage system
  
  • Ruty Guillaume
  • Surcouf André
  • Rougier Jean-Louis
  , 2017. The exponentially growing storage demands that industry currently faces puts a huge stress on traditional distributed storage systems. Indeed, the concurrently increasing quantity of available video content and video quality comes with a new set of challenges that existing distributed systems have a hard time to meet. These difficulties are mostly the result of design decisions that, while relevant for the initial scope of those systems, end up limiting their scalability. In this paper, a quick overview of these typical design decisions is presented for two examples: Ceph and GFS (Google File System). An original and IPv6-centric architecture is then described, that presents none of the aforementioned scalability limitations, and a first prototype of this architecture is compared to Ceph. (10.1109/SDS.2017.7939145)
  DOI : 10.1109/SDS.2017.7939145
• Proxy Clouds for RGB-D Stream Processing: A preview
  
  • Kaiser Adrien
  • Ybanez Zepeda Jose Alonso
  • Boubekeur Tamy
  , 2017.
• Subcarrier and Power allocation for Sum Rate Maximization in Multicarrier NOMA Systems
  
  • Fu Yaru
  • Salaun Lou
  • Sung Chi Wan
  • Chen Chung Shue Calvin
  • Coupechoux Marceau
  , 2017. This paper investigates the subcarrier and power allocation for the downlink of a multicarrier non-orthogonal multiple access (MC-NOMA) system. A three-step algorithm is proposed to deal with the sum rate maximization problem. First, we assume each user could use all the subcarriers si- multaneously and apply the synchronous iterative waterfilling algorithm (SIWA) to obtain a power vector for each user. Second, subcarriers are assigned to users by a heuristic greedy method based on the achieved power allocation result of Step 1. Third, SIWA is used once again to further improve the system performance with the obtained subcarrier assignment result of Step 2. In Step 3, where the number of multiplexed users is no more than two, the convergence of SIWA is proved analytically. Numerical results show that the proposed resource allocation algorithm achieves comparable performance to an existing near- optimal solution but with much lower time complexity.
• Conception d’étalons coplanaires couplées pour la méthode d’étalonnage Multimode TRL
  
  • Thi-Dao Pham
  • Allal Djamel
  • Ziadé F.
  • Bergeault Eric
  , 2017.
• Sparse Stochastic Bandits
  
  • Kwon Joon
  • Perchet Vianney
  • Vernade Claire
  , 2017.
• Séries SAR multi-temporelles et applications en milieu urbain
  
  • Tupin Florence
  , 2017.
• Calibration des antennes actives avec prise en compte des effets de bord et de couplage
  
  • Chalumyan Taguhi
  • Maas Olivier
  • Begaud Xavier
  • Huyart Bernard
  , 2017.
• Absorbant Ultra Large Bande sur un large secteur d’angles d’incidence
  
  • Begaud Xavier
  • Varault Stefan
  • Lepage A. C.
  • Soiron M.
  • Barka André
  , 2017.
• Energy Efficient Magnetic Tunnel Junction Based Hybrid LSI Using Multi-Threshold UTBB-FD-SOI Device
  
  • Cai Hao
  • Wang You
  • Naviner Lirida
  • Zhao Weisheng
  , 2017.
• Path separation by short cycles
  
  • Cohen Gérard
  • Fachini Emanuela
  • Korner J.
  Journal of Graph Theory, 2017, 85 (1), pp.107-114.
• Sharp Bernstein and Hoeffding type inequalities for regenerative Markov chains
  
  • Ciolek Gabriela
  • Bertail Patrice
  , 2017.
• Internet of Things (IoT) Analytics
  
  • Bifet Albert
  , 2017.
• Analysis of the influence of diversity in collaborative and multi-view clustering
  
  • Sublime Jérémie
  • Matei Basarab
  • Murena Pierre-Alexandre
  , 2017. Multi-source clustering is common data mining task the aim of which is to use several clustering algorithms to analyze different aspects of the same data. Well known applications of multi-source clustering include horizontal collaborative clustering and multi-view clustering, where several algorithms combine their strengths by exchanging information about their finding on local structures with a goal of mutual improvement. However, many of these proposed algorithms and statistical models lack the capability to detect weak collaborations that may prove detrimental to the global clustering process. In this article, we propose a weighing optimization method that will help detecting which algorithms should exchange their information based on the diversity between the different algorithms’ solutions.
• Bootstrapping periodically autoregressive models
  
  • Ciolek Gabriela
  , 2017.
• Frequency-Dependence of Channel Delay Spread in an Outdoor Environment
  
  • Diakhate Cheikh
  • Conrat Jean-Marc
  • Cousin Jean-Christophe
  • Sibille Alain
  , 2017. Millimeter-Wave (mmW) communication systems, capable of achieving high data rates thanks to the large bandwidth available in this frequency range, are a promising 5G technology. Studies in this paper investigate the radio propagation channel at 3, 17 and 60 GHz in an Outdoor environment. Measurements were conducted using a wideband channel sounder to derive channel delay spread (DS). Results do not indicate a clear trend on the behavior of the DS with regards to the frequency. This disagrees in some way with the 3GPP TR 38.900 channel model findings which support a clear decreasing trend of this parameter as the frequency increases. At 60 GHz, the channel DS was computed with and without compensation of atmospheric oxygen absorption. It was found that this phenomenon has minor impact on the measurement results. Its compensation does not change the main conclusions.
• The local-global principle for real analysis
  
  • Rioul Olivier
  , 2017. The logical foundations of real analysis were developed in the end of 19th century and beginning of 20th century by mathematicians such as Bolzano, Cauchy, Weierstrass, Dedekind, Cantor, Heine, Borel, Cousin and Lebesgue. Motivated by teaching, they departed from the geometric intuition of the “real line” by establishing rigorous proofs based on “completeness” axioms that char- acterize the real number continuum. Today, the foundation is still recognized as satisfactory, and all classical textbooks define R as any ordered field sat- isfying one of the many equivalent completeness axioms, such as the l.u.b. property, Dedekind cuts, Cantor’s property of nested intervals, monotone con- vergence, Bolzano-Weierstrass, Borel-Lebesgue, Cousin’s partitions, continuous induction, etc. It is somewhat striking that all these equivalent axioms look so diverse. Also, several classical proofs of the basic theorems in real analysis seem difficult and subtle for the beginner (e.g., proofs of the extreme value theorem or Heine’s the- orem using Bolzano-Weierstrass). This calls for a need of a unifying principle from which all other axioms and basic continuity theorems could be easily de- rived. We discuss yet another equivalent axiom in two equivalent versions of the form: Local-Global: Any local and additive property is global; Global-Local: Any global and subtractive property has a limit point. The earliest reference that explicitly describes this principle is Guyou’s little- known French textbook [1]. It was later re-discovered independently many times in many disguises in some American circles [2, 3, 4, 5, 6]. The modern presen- tation we give is both enjoyable and successful for obtaining easy proofs of all useful theorems of real analysis, including those for continuity and differentia- bility of functions of the real variable. As a perspective it can be extended to a new presentation of the integral (just as powerful as the Kurzweil-Henstock integral—which is more powerful then Lebesgue’s integral) where the so-called “fundamental theorem of calculus” becomes the actual definition for a general- ized notion of the antiderivative (primitive) function. This work was initiated from a joint collaboration with José Carlos Magossi, Faculdade de Tecnologia, Universidade Estadual de Campinas, Brazil.
• Data stream classification using random feature functions and novel method combinations
  
  • Marrón Diego
  • Read Jesse
  • Bifet Albert
  • Navarro Nacho
  Journal of Systems and Software, Elsevier, 2017, 127, pp.195 - 204. (10.1016/j.jss.2016.06.009)
  DOI : 10.1016/j.jss.2016.06.009
• Impact of the switching activity on the aging of delay-PUFs
  
  • Karimi Naghmeh
  • Danger Jean-Luc
  • Lozac'H Florent
  • Guilley Sylvain
  , 2017. (10.1109/ETS.2017.7968223)
  DOI : 10.1109/ETS.2017.7968223
• Concentration inequalities for regenerative and Harris recurrent Markov chains with applications to statistical learning
  
  • Clémençon Stéphan
  • Ciolek Gabriela
  • Bertail P.
  , 2017. Concentration inequalities are very often a crucial step in deriving many results in statistical learning. The purpose of this talk is to present Bernstein and Hoeffding type maximal inequalities for regenerative Markov chains. Furthermore, we generalize these results and show exponential bounds for suprema of empirical processes over a class of functions F which size is controlled by its uniform entropy number. We show also that concentration inequalities are possible to obtain when the chain is sub-geometric. All constants involved in the bounds of the considered inequalities are given in an explicit form which can be advantageous in practical considerations. We show that the inequalities obtained for regenerative Markov chains can be easily generalized to a Harris recurrent case. Finally we provide one example of application of presented inequalities in statistical learning theory and obtain generalization bounds for mimimum volume set estimation problem when the data are Markovian.
• Accelerated encryption algorithms for secure storage and processing in the cloud
  
  • Badii Atta
  • Faulkner Ryan
  • Glackin Cornelius
  • Chollet Gérard
  , 2017, pp.1-6. The objective of this paper is to outline the design specification, implementation and evaluation of a proposed accelerated encryption framework which deploys both homomorphic and symmetric-key encryptions to serve the privacy preserving processing; in particular, as a sub-system within the Privacy Preserving Speech Processing framework architecture as part of the PPSP-in-Cloud Platform. Following a preliminary study of GPU efficiency gains optimisations benchmarked for AES implementation we have addressed and resolved the Big Integer processing challenges in parallel implementation of bilinear pairing thus enabling the creation of partially homomorphic encryption schemes which facilitates applications such as speech processing in the encrypted domain on the cloud. This novel implementation has been validated in laboratory tests using a standard speech corpus and can be used for other …
• Optimal distributed channel assignment in D2D networks using learning in noisy potential games
  
  • Ali Mohammed Shabbir
  • Coucheney Pierre
  • Coupechoux Marceau
  , 2017, pp.151-156. (10.1109/INFCOMW.2017.8116368)
  DOI : 10.1109/INFCOMW.2017.8116368
• Superstrat tout diélectrique pour le contrôle du rayonnement d’une antenne – Prototypage et mesures
  
  • Joshi Chetan
  • Lepage A. C.
  • Begaud Xavier
  • Piau Gérard-Pascal
  , 2017.