Publications

2019

• Skin-On Interfaces: A Bio-Driven Approach for Artificial Skin Design to Cover Interactive Devices
  
  • Teyssier Marc
  • Bailly Gilles
  • Pelachaud Catherine
  • Lecolinet Eric
  • Conn Andrew R.
  • Roudaut Anne
  , 2019, pp.307-322. We propose a paradigm called Skin-On interfaces, in which interactive devices have their own (artificial) skin, thus enabling new forms of input gestures for end-users (e.g. twist, scratch). Our work explores the design space of Skin-On interfaces by following a bio-driven approach: (1) From a sensory point of view, we study how to reproduce the look and feel of the human skin through three user studies; (2) From a gestural point of view, we explore how gestures naturally performed on skin can be transposed to Skin-On interfaces; (3) From a technical point of view, we explore and discuss different ways of fabricating interfaces that mimic human skin sensitivity and can recognize the gestures observed in the previous study; (4) We assemble the insights of our three exploratory facets to implement a series of Skin-On interfaces and we also contribute by providing a toolkit that enables easy reproduction and fabrication. (10.1145/3332165.3347943)
  DOI : 10.1145/3332165.3347943
• IDENTIFY, LOCATE AND SEPARATE: AUDIO-VISUAL OBJECT EXTRACTION IN LARGE VIDEO COLLECTIONS USING WEAK SUPERVISION
  
  • Parekh Sanjeel
  • Ozerov Alexey
  • Essid Slim
  • Duong Ngoc
  • Pérez Patrick
  • Richard Gael
  , 2019.
• Weakly informed audio source separation
  
  • Schulze-Forster Kilian
  • Doire Clément
  • Richard Gael
  • Badeau Roland
  , 2019.
• Regular graphs from weakly regular plateaued functions
  
  • Mesnager Sihem
  • Sinak A.
  , 2019.
• Nonlinear Functions in Learned Iterative Shrinkage-Thresholding Algorithm for Sparse Signal Recovery
  
  • Marques Elaine Crespo
  • Maciel Nilson
  • Naviner Lirida
  • Cai Hao
  • Yang Jun
  , 2019, pp.324-329. Compressive sensing requires fewer measurements than Nyquist rate to recover sparse signals, leading to processing and energy saving. The efficiency of this technique strongly depends on the quality of the considered sparse recovery algorithm. This work focuses on a learned iterative shrinkage-thresholding algorithm where iterations are related to layers of a neural network. We analyze the performance of this algorithm for different shrinkage functions. A decrease up to 9dB in the NMSE value is achieved by choosing appropriate shrinkage function. Moreover, the estimation performance can be close to the theoretical performance bound, showing deep learning as a promising tool for sparse signal estimation. This work can be applied in several areas such as image processing, Internet of Things (IoT), cognitive radio networks, and sparse channel estimation for wireless communications. (10.1109/SiPS47522.2019.9020469)
  DOI : 10.1109/SiPS47522.2019.9020469
• MNE: Software for Acquiring, Processing, and Visualizing MEG/EEG Data
  
  • Esch Lorenz
  • Dinh Christoph
  • Larson Eric
  • Engemann Denis
  • Jas Mainak
  • Khan Sheraz
  • Gramfort Alexandre
  • Hämäläinen Matti
  , 2019, pp.355-371. The methods for acquiring, processing, and visualizing magnetoencephalography (MEG) and electroencephalography (EEG) data are rapidly evolving. Advancements in hardware and software development offer new opportunities for cognitive and clinical neuroscientists but at the same time introduce new challenges as well. In recent years the MEG/EEG community has developed a variety of software tools to overcome these challenges and cater to individual research needs. As part of this endeavor, the MNE software project, which includes MNE-C, MNE-Python, MNE-CPP, and MNE-MATLAB as its subprojects, offers an efficient set of tools addressing certain common needs. Even more importantly, the MNE software family covers diverse use case scenarios. Here, we present the landscape of the MNE project and discuss how it will evolve to address the current and emerging needs of the MEG/EEG community. (10.1007/978-3-030-00087-5_59)
  DOI : 10.1007/978-3-030-00087-5_59
• Fast and robust modelling using a direct translation from a robotic application to its abstracted behaviour
  
  • Rataj Artur
  • Borde Etienne
  , 2019, pp.50-56. (10.1145/3339985.3358492)
  DOI : 10.1145/3339985.3358492
• Automated Spinal Midline Delineation on Biplanar X-Rays Using Mask R-CNN
  
  • Yang Zixin
  • Skalli Wafa
  • Vergari Claudio
  • Angelini Elsa
  • Gajny Laurent
  , 2019, 34, pp.307-316. Manually annotating medical images with few landmarks to initialize 3D shape models is a common practice. For instance, when reconstructing the 3D spine from biplanar X-rays, the spinal midline, passing through vertebrae body centers (VBCs) and endplate midpoints, is required. This paper presents an automated spinal midline delineation method on frontal and sagittal views by using Mask R-CNN. The network detects all vertebrae from C7 to L5, followed by vertebrae segmentation and classification at the same time. After postprocessing to discard outliers, the vertebrae mask centers were regarded as VBCs to get the spine midline by polynomial fitting. Evaluation of the spinal midline on 136 images used root mean square error (RMSE) with respect to manual ground-truth. The RMSE ± standard error values of predicted spinal midlines (C7-L5) were 1.11 mm ± 0.67 mm on frontal views and 1.92mm ± 1.38 mm on sagittal views. The proposed method is capable of delineating spinal midlines on patients with different spine deformity degrees. (10.1007/978-3-030-32040-9_32)
  DOI : 10.1007/978-3-030-32040-9_32
• Appearance and Pose-Conditioned Human Image Generation using Deformable GANs
  
  • Siarohin Aliaksandr
  • Lathuilière Stéphane
  • Sangineto Enver
  • Sebe Nicu
  IEEE Transactions on Pattern Analysis and Machine Intelligence, Institute of Electrical and Electronics Engineers, 2019. In this paper, we address the problem of generating person images conditioned on both pose and appearance information. Specifically, given an image xa of a person and a target pose P(xb), extracted from a different image xb, we synthesize a new image of that person in pose P(xb), while preserving the visual details in xa. In order to deal with pixel-to-pixel misalignments caused by the pose differences between P(xa) and P(xb), we introduce deformable skip connections in the generator of our Generative Adversarial Network. Moreover, a nearest-neighbour loss is proposed instead of the common L1 and L2 losses in order to match the details of the generated image with the target image. Quantitative and qualitative results, using common datasets and protocols recently proposed for this task, show that our approach is competitive with respect to the state of the art. Moreover, we conduct an extensive evaluation using off-the-shell person re-identification (Re-ID) systems trained with person-generation based augmented data, which is one of the main important applications for this task. Our experiments show that our Deformable GANs can significantly boost the Re-ID accuracy and are even better than data-augmentation methods specifically trained using Re-ID losses.
• VisualTouch: Enhancing Affective Touch Communication with Multi-modality Stimulation
  
  • Zhang Zhuoming
  • Heron Robin
  • Lecolinet Eric
  • Detienne Françoise
  • Safin Stéphane
  , 2019. As one of the most important non-verbal communication channel, touch plays an essential role in interpersonal af-fective communication. Although some researchers have started exploring the possibility of using wearable devices for conveying emotional information, most of the existing devices still lack the capability to support affective and dynamic touch in interaction. In this paper, we explore the effect of dynamic visual cues on the emotional perception of vibrotactile signals. For this purpose, we developed Visu-alTouch, a haptic sleeve consisting of a haptic layer and a visual layer. We hypothesized that visual cues would enhance the interpretation of tactile cues when both types of cues are congruent. We first carried out an experiment and selected 4 stimuli producing substantially different responses. Based on that, a second experiment was conducted with 12 participants rating the valence and arousal of 36 stimuli using SAM scales. (10.1145/3340555.3353733)
  DOI : 10.1145/3340555.3353733
• Scalable Byzantine Reliable Broadcast
  
  • Guerraoui Rachid
  • Kuznetsov Petr
  • Monti Matteo
  • Pavlovic Matej
  • Seredinschi Dragos-Adrian
  , 2019.
• Adaptive optics assisted space-ground coherent optical links: ground receiver performance with digital phase locked loop
  
  • Paillier Laurie
  • Conan Jean-Marc
  • Le Bidan Raphaël
  • Artaud Géraldine
  • Védrenne Nicolas
  • Jaouën Yves
  , 2019. Combination of phase modulation with coherent detection for low Earth orbit satellite-to-ground optical links is currently investigated to meet the current high-data-rate increasing demand. On the downlink, the signal carrier undergoes a large Doppler frequency shift which may in turn severely hinders demodulation and information recovery. We present a coherent receiver architecture which combines adaptive optics correction, to mitigate the atmospheric turbulence detrimental effect, with a digital carrier synchronization system. The latter is based on a phase-locked loop which ensures the compensation of the residual frequency shift that remains after preliminary coarse frequency offset correction. We describe the methodology followed to design such a digital phase-locked loop. As an example, we show that the digital PLL is able to compensate a residual frequency mismatch of 300MHz with a convergence time inferior to 1s which is reasonable compared to the duration of the satellite pass. (10.1109/ICSOS45490.2019.8978983)
  DOI : 10.1109/ICSOS45490.2019.8978983
• Connectivity-preserving Smooth Surface Filling with Sharp Features
  
  • Lescoat Thibault
  • Memari Pooran
  • Thiery Jean-Marc
  • Ovsjanikov Maks
  • Boubekeur Tamy
  , 2019. We present a method for constructing a surface mesh filling gaps between the boundaries of multiple disconnected input components. Unlike previous works, our method pays special attention to preserving both the connectivity and large-scale geometric features of input parts, while maintaining efficiency and scalability w.r.t. mesh complexity. Starting from an implicit surface reconstruction matching the parts' boundaries, we first introduce a modified dual contouring algorithm which stitches a meshed contour to the input components while preserving their connectivity. We then show how to deform the reconstructed mesh to respect the boundary geometry and preserve sharp feature lines, smoothly blending them when necessary. As a result, our reconstructed surface is smooth and propagates the feature lines of the input. We demonstrate on a wide variety of input shapes that our method is scalable to large input complexity and results in superior mesh quality compared to existing techniques. (10.2312/pg.20191332)
  DOI : 10.2312/pg.20191332
• Fast and Scalable Optimal Transport for Brain Tractograms
  
  • Feydy Jean
  • Roussillon Pierre
  • Trouvé Alain
  • Gori Pietro
  , 2019. We present a new multiscale algorithm for solving regular-ized Optimal Transport problems on the GPU, with a linear memory footprint. Relying on Sinkhorn divergences which are convex, smooth and positive definite loss functions, this method enables the computation of transport plans between millions of points in a matter of minutes. We show the effectiveness of this approach on brain tractograms modeled either as bundles of fibers or as track density maps. We use the resulting smooth assignments to perform label transfer for atlas-based segmentation of fiber tractograms. The parameters-blur and reach-of our method are meaningful, defining the minimum and maximum distance at which two fibers are compared with each other. They can be set according to anatomical knowledge. Furthermore, we also propose to estimate a probabilistic atlas of a population of track density maps as a Wasserstein barycenter. Our CUDA implementation is endowed with a user-friendly PyTorch interface, freely available on the PyPi repository (pip install geomloss) and at www.kernel-operations.io/geomloss. (10.1007/978-3-030-32248-9_71)
  DOI : 10.1007/978-3-030-32248-9_71
• DC Coefficients Recovery from AC Coefficients in the JPEG Compression Scenario
  
  • Qiu Han
  • Zheng Qinkai
  • Qiu Meikang
  • Memmi Gérard
  , 2019, pp.266-276. (10.1007/978-3-030-34139-8_26)
  DOI : 10.1007/978-3-030-34139-8_26
• Cloud-Radio Access Networks : design, optimization and algorithms
  
  • Mharsi Niezi
  , 2019. Cloud Radio Access Network (C-RAN) has been proposed as a promising architecture to meet the exponential growth in data traffic demands and to overcome the challenges of next generation mobile networks (5G). The main concept of C-RAN is to decouple the BaseBand Units (BBU) and the Remote Radio Heads (RRH), and place the BBUs in common edge data centers (BBU pools) for centralized processing. This gives a number of benefits in terms of cost savings, network capacity improvement and resource utilization gains. However, network operators need to investigate scalable and cost-efficient algorithms for resource allocation problems to enable and facilitate the deployment of C-RAN architecture. Most of these problems are very complex and thus very hard to solve. Hence, we use combinatorial optimization which provides powerful tools to efficiently address these problems.One of the key issues in the deployment of C-RAN is finding the optimal assignment of RRHs (or antennas) to edge data centers (BBUs) when jointly optimizing the fronthaul latency and resource consumption. We model this problem by a mathematical formulation based on an Integer Linear Programming (ILP) approach to provide the optimal strategies for the RRH-BBU assignment problem and we propose also low-complexity heuristic algorithms to rapidly reach good solutions for large problem instances. The optimal RRH-BBU assignment reduces the expected latency and offers resource utilization gains. Such gains can only be achieved when reducing the inter-cell interference caused by the dense deployment of cell sites. We propose an exact mathematical formulation based on Branch-and-Cut methods that enables to consolidate and re-optimize the antennas radii in order to jointly minimize inter-cell interference and guarantee a full network coverage in C-RAN. In addition to the increase of inter-cell interference, the high density of cells in C-RAN increases the amount of baseband processing as well as the amount of data traffic demands between antennas and centralized data centers when strong latency requirements on fronthaul network should be met. Therefore, we discuss in the third part of this thesis how to determine the optimal placement of BBU functions when considering 3GPP split option to find optimal tradeoffs between benefits of centralization in C-RAN and transport requirements. We propose exact and heuristic algorithms based on combinatorial optimization techniques to rapidly provide optimal or near-optimal solutions even for large network sizes.
• A Misbehavior Authority System for Sybil Attack Detection in C-ITS
  
  • Kamel Joseph
  • Haidar Farah
  • Jemaa Ines Ben
  • Kaiser Arnaud
  • Lonc Brigitte
  • Urien Pascal
  , 2019. Global misbehavior detection is an important back-end mechanism in Cooperative Intelligent Transport Systems (C-ITS). It is based on the local misbehavior detection information sent by Vehicle's On-Board Units (OBUs) and by RoadSide Units (RSUs) called Misbehavior Reports (MBRs) to the Mis-behavior Authority (MA). By analyzing these reports, the MA provides more accurate and robust misbehavior detection results. Sybil attacks pose a significant threat to the C-ITS systems. Their detection and identification may be inaccurate and confusing. In this work, we propose a Machine Learning (ML) based solution for the internal detection process of the MA. We show through extensive simulation that our solution is able to precisely identify the type of the Sybil attack and provide promising detection accuracy results.
• 3D Simulation to Validate Autonomous Intervention Systems Architecture for Disaster Management
  
  • Tanzi Tullio
  • Bertolino Matteo
  , 2020, AICT-575, pp.196-211. The use of autonomous robots either on the ground (i.e., Rover) or flying (i.e., Drone) constitutes a major progress in the support of a crisis. To work properly and to reach the desired level of autonomy, they have to be correctly configured though. Indeed, errors on robot configuration can lead to imprecise or erroneous data and, consequently, erroneous decisions can result from them. Before the beginning of the mission, it is important also to achieve a strong level of confidence about the usage of the sensors (for example, LIDARs) with respect to the context of the mission. Many aspects of these validations cannot be performed during the mission, for example verifying the behaviour of a rover following a strong collision with an external actor (such as debris) that can potentially damage or break some components. Moreover, during a real mission it is not always possible making huge modifications in the system configuration. In this respect, simulating the behaviour of the system in a virtual environment, similar to the real physical world, can constitute a good validation approach before the mission. These simulations allow to validate the behaviour and the configuration of the system as well as the most appropriate equipment of it. (10.1007/978-3-030-48939-7_17)
  DOI : 10.1007/978-3-030-48939-7_17
• Towards 3D Simulation to Validate Autonomous Systems Intervention in Disaster Management Environment
  
  • Tanzi Tullio Joseph
  • Bertolino Matteo
  , 2019.
• Programmable Optical x-Haul Network in the COSMOS Testbed
  
  • Gutterman Craig
  • Minakhmetov Artur
  • Yu Jiakai
  • Sherman Michael
  • Chen Tingjun
  • Zhu Shengxiang
  • Seskar Ivan
  • Raychaudhuri Dipankar
  • Kilper Daniel
  • Zussman Gil
  , 2019, pp.1-2. (10.1109/ICNP.2019.8888108)
  DOI : 10.1109/ICNP.2019.8888108
• Improved small molecule identification through learning combinations of kernel regression models
  
  • Brouard Celine
  • d'Alché-Buc Florence
  • Rousu Juho
  , 2019.
• An efficient key distribution system for data fusion in V2X heterogeneous networks
  
  • Qiu Han
  • Qiu Meikang
  • Lu Zhihui
  • Memmi Gérard
  Information Fusion, Elsevier, 2019, 50, pp.212-220. (10.1016/j.inffus.2019.02.002)
  DOI : 10.1016/j.inffus.2019.02.002
• Complexity of the Uniqueness of an Optimal Identifying or Locating-Dominating Code
  
  • Hudry Olivier
  , 2019.
• All-Or-Nothing data protection for ubiquitous communication: Challenges and perspectives
  
  • Qiu Han
  • Kapusta Katarzyna
  • Lu Zhihui
  • Qiu Meikang
  • Memmi Gérard
  Information Sciences, Elsevier, 2019, 502, pp.434-445. (10.1016/j.ins.2019.06.031)
  DOI : 10.1016/j.ins.2019.06.031
• High frequency dynamics in quantum cascade lasers : a roadmap to free-space communications in the mid-infrared
  
  • Spitz O
  • Herdt A
  • Maisons G.
  • Carras M.
  • Elsässer W
  • Grillot F.
  , 2019. Quantum cascade lasers, which can emit deterministic chaotic patterns, are found to exhibit improved chaos properties when using optical injection instead of feedback. These findings pave a way for high-speed secure communications in the mid-infrared