Publications

2020

• PolySense: Augmenting Textiles with Electrical Functionality using In-Situ Polymerization
  
  • Honnet Cedric
  • Perner-Wilson Hannah
  • Teyssier Marc
  • Fruchard Bruno
  • Steimle Jürgen
  • Baptista Ana C
  • Strohmeier Paul
  , 2020, pp.3376841. We present a method for enabling arbitrary textiles to sense pressure and deformation: In-situ polymerization supports integration of piezoresistive properties at the material level, preserving a textile's haptic and mechanical characteristics. We demonstrate how to enhance a wide set of fabrics and yarns using only readily available tools. To further support customisation by the designer, we present methods for patterning, as needed to create circuits and sensors, and demonstrate how to combine areas of different conductance in one material. Technical evaluation results demonstrate the performance of sensors created using our method is comparable to off-the-shelf piezoresistive textiles. As application examples, we demonstrate rapid manufacturing of on-body interfaces, tie-dyed motion-capture clothing, and zippers that act as potentiometers. (10.1145/3313831.3376841)
  DOI : 10.1145/3313831.3376841
• Understanding the Heisenberg Effect of Spatial Interaction: A Selection Induced Error for Spatially Tracked Input Devices
  
  • Wolf Dennis
  • Gugenheimer Jan
  • Combosch Marco
  • Rukzio Enrico
  , 2020, pp.1-10. (10.1145/3313831.3376876)
  DOI : 10.1145/3313831.3376876
• Towards Inclusive External Communication of Autonomous Vehicles for Pedestrians with Vision Impairments
  
  • Colley Mark
  • Walch Marcel
  • Gugenheimer Jan
  • Askari Ali
  • Rukzio Enrico
  , 2020, pp.1-14. (10.1145/3313831.3376472)
  DOI : 10.1145/3313831.3376472
• Exploring Potentially Abusive Ethical, Social and Political Implications of Mixed Reality Research in HCI
  
  • Gugenheimer Jan
  • Mcgill Mark
  • Huron Samuel
  • Mai Christian
  • Williamson Julie
  • Nebeling Michael
  , 2020, pp.1-8. (10.1145/3334480.3375180)
  DOI : 10.1145/3334480.3375180
• VRSketchIn:Exploring the Design Space of Pen and Tablet Interaction for 3D Sketching in Virtual Reality
  
  • Drey Tobias
  • Gugenheimer Jan
  • Karlbauer Julian
  • Milo Maximilian
  • Rukzio Enrico
  , 2020, pp.1-14. (10.1145/3313831.3376628)
  DOI : 10.1145/3313831.3376628
• Mix&Match: Towards Omitting Modelling Through In-situ Remixing of Model Repository Artifacts in Mixed Reality
  
  • Stemasov Evgeny
  • Wagner Tobias
  • Gugenheimer Jan
  • Rukzio Enrico
  , 2020, pp.1-12. (10.1145/3313831.3376839)
  DOI : 10.1145/3313831.3376839
• confStream: Automated Algorithm Selection and Configuration of Stream Clustering Algorithms
  
  • Carnein Matthias
  • Trautmann Heike
  • Bifet Albert
  • Pfahringer Bernhard
  , 2020, 12096, pp.80--95. Machine learning has become one of the most important tools in data analysis. However, selecting the most appropriate machine learning algorithm and tuning its hyperparameters to their optimal values remains a difficult task. This is even more difficult for streaming applications where automated approaches are often not available to help during algorithm selection and configuration. This paper proposes the first approach for automated algorithm selection and configuration of stream clustering algorithms. We train an ensemble of different stream clustering algorithms and configurations in parallel and use the best performing configuration to obtain a clustering solution. By drawing new configurations from better performing ones, we are able to improve the ensemble performance over time. In large experiments on real and artificial data we show how our ensemble approach can improve upon default configurations and can also compete with a-posteriori algorithm configuration. Our approach is considerably faster than a-posteriori approaches and applicable in real-time. In addition, it is not limited to stream clustering and can be generalised to all streaming applications, including stream classification and regression. (10.1007/978-3-030-53552-0_10)
  DOI : 10.1007/978-3-030-53552-0_10
• Semantic feature selection for network telemetry event description
  
  • Feltin Thomas
  • Foroughi Parisa
  • Shao Wenqin
  • Brockners Frank
  • Clausen Thomas Heide
  , 2020, pp.1-6. Model driven telemetry (MDT) enables the real-time collection of hundreds of thousands of counters on large-scale networks, with contextual information to each counter provided in the telemetry data structure definition. Explaining network events in such datasets implies substantial analysis by a domain expert. This paper presents an semantic feature selection method, to find the most important counters which describe a given event in a telemetry dataset, and facilitate the explanation process. This paper proposes a metric for estimating the importance of features in a dataset with descriptive feature names, to find those that are most meaningful to a human. With this estimation, this paper presents a cross-entropy based metric describing the quality of a selection of counters, which is combined with the data behavior to define an optimization goal. The computation of optimal selections distills intelligible and precise selections of counters with adjustable verbosity, and describes events with a few selected counters outlining the root cause of network events. (10.1109/NOMS47738.2020.9110382)
  DOI : 10.1109/NOMS47738.2020.9110382
• Detection of Side-channel Lleakage Through Glitches Using an Automated Tool
  
  • Sauvage Laurent
  • Takarabt Sofiane
  • Souissi Youssef
  • Guilley Sylvain
  • Mathieu Yves
  , 2020.
• Exact solutions and analysis of an SIR variant with constant-time recovery
  
  • Madore David Alexander
  , 2020. We investigate a variant of the SIR epidemiological model in which the recovery of infected individuals takes place in constant time rather than following an exponential distribution. This model is described by a delay-differential equation: we show that the equations in question admit an exact solution in closed form (given by rational functions of an exponential of time). Using this, we investigate the qualitative differences between this modified model and classical SIR and show that, for the same reproduction number, contagiousness and expected recovery time, the constant-time recovery variant entails a sharper, more pronounced, epidemiological peak than the classical variant (exponential-process recovery), while still having the same final attack rate.
• Language: The missing selection pressure
  
  • Dessalles Jean-Louis
  Theoria et Historia Scientiarum, 2020, 17, pp.1. Human beings are talkative. What advantage did their ancestors find in communicating so much? Numerous authors consider this advantage to be “obvious” and “enormous”. If so, the problem of the evolutionary emergence of language amounts to explaining why none of the other primate species evolved anything even remotely similar to language. I propose to reverse the picture. On closer examination, language resembles a losing strategy. Competing for providing other individuals with information, sometimes striving to be heard, makes apparently no sense within a Darwinian framework. At face value, language as we can observe it should never have existed or should have been counter-selected. In other words, the selection pressure that led to language is still missing. The solution I propose consists in regarding language as a social signaling device that developed in a context of generalized insecurity that is unique to our species. By talking, individuals advertise their alertness and their ability to get informed. This hypothesis is shown to be compatible with many characteristics of language that otherwise are left unexplained. (10.12775/ths.2020.001)
  DOI : 10.12775/ths.2020.001
• Systematic investigation of the influencing parameters of an external cavity laser with a quantum dot gain chip
  
  • Ehlert Jannik F.
  • Mugnier Alain
  • He Gang
  • Grillot Frederic
  , 2020, 11356 (9). External cavity lasers show a variety of uses, for which quantum well semiconductor lasers are already commercially used. Due to the atom-like discrete energy levels, quantum dots exhibit various properties resulting from the three-dimensional confinement of carriers, like high stability against temperature variation, large gain bandwidth, and low-threshold lasing operation. Quantum dots seem to be ideal to address the challenges in the further development of various semiconductor applications, such as high-resolution spectroscopy or broad-band optical communication networks, for which a range of spectral and temporal characteristics is required, for instance a narrow spectral linewidth, low intensity noise or wide wavelength tunability. In this view, external cavity quantum dot gain chips can be envisoned to replace the current quantum well technology. Using a semi-analytical rate equation model, we successfully analyze both dynamical and noise properties of an external cavity laser made with quantum dot gain medium, operating under strong optical feedback. This paper investigates the turn-on delay, the relative intensity noise, and the frequency noise and compares them to the case without optical feedback. These numerical investigations of an external cavity quantum dot gain chip provide meaningful building blocks for future fabrication research or for developing high performance device such as wavelength-selective components. (10.1117/12.2554553)
  DOI : 10.1117/12.2554553
• Reconstruction for Diverging-Wave Imaging Using Deep Convolutional Neural Networks
  
  • Lu Jingfeng
  • Millioz Fabien
  • Garcia Damien
  • Salles Sébastien
  • Liu Wanyu
  • Friboulet Denis
  IEEE Transactions on Ultrasonics, Ferroelectrics and Frequency Control, Institute of Electrical and Electronics Engineers, 2020, 67 (12), pp.2481-2492. (10.1109/TUFFC.2020.2986166)
  DOI : 10.1109/TUFFC.2020.2986166
• Growing Random Graphs with Quantum Rules
  
  • Jnane Hamza
  • Giuseppe Di Molfetta
  • Miatto Filippo
  Electronic Proceedings in Theoretical Computer Science, EPTCS, 2020, 315, pp.38-47. Random graphs are a central element of the study of complex dynamical networks such as the internet, the brain, or socioeconomic phenomena. New methods to generate random graphs can spawn new applications and give insights into more established techniques. We propose two variations of a model to grow random graphs and trees, based on continuous-time quantum walks on the graphs. After a random characteristic time, the position of the walker(s) is measured and new nodes are attached to the nodes where the walkers collapsed. Such dynamical systems are reminiscent of the class of spontaneous collapse theories in quantum mechanics. We investigate several rates of this spontaneous collapse for an individual quantum walker and for two non-interacting walkers. We conjecture (and report some numerical evidence) that the models are scale-free. (10.4204/EPTCS.315.4)
  DOI : 10.4204/EPTCS.315.4
• Guiding Attention in Sequence-to-Sequence Models for Dialogue Act Prediction
  
  • Colombo Pierre
  • Chapuis Emile
  • Manica Matteo
  • Vignon Emmanuel
  • Varni Giovanna
  • Clavel Chloé
  Proceedings of the AAAI Conference on Artificial Intelligence, Association for the Advancement of Artificial Intelligence, 2020, 34 (05), pp.7594-7601. The task of predicting dialog acts (DA) based on conversational dialog is a key component in the development of conversational agents. Accurately predicting DAs requires a precise modeling of both the conversation and the global tag dependencies. We leverage seq2seq approaches widely adopted in Neural Machine Translation (NMT) to improve the modelling of tag sequentiality. Seq2seq models are known to learn complex global dependencies while currently proposed approaches using linear conditional random fields (CRF) only model local tag dependencies. In this work, we introduce a seq2seq model tailored for DA classification using: a hierarchical encoder, a novel guided attention mechanism and beam search applied to both training and inference. Compared to the state of the art our model does not require handcrafted features and is trained end-to-end. Furthermore, the proposed approach achieves an unmatched accuracy score of 85% on SwDA, and state-of-the-art accuracy score of 91.6% on MRDA. (10.1609/aaai.v34i05.6259)
  DOI : 10.1609/aaai.v34i05.6259
• Co-Engineering Gap Analysis of ANSI/ISA‑62443‑3‑3
  
  • Mlynek Petr
  • Fujdiak Radek
  • Mrnustik Pavel
  • Krena Bohuslav
  • Apvrille Ludovic
  International journal of advances in telecommunications, electrotechnics, signals and systems, International science and engineering society, 2020, 9 (1), pp.1. (10.11601/ijates.v9i1.285)
  DOI : 10.11601/ijates.v9i1.285
• High Throughput/Gate AES Hardware Architectures Based on Datapath Compression
  
  • Ueno Rei
  • Homma Naofumi
  • Morioka Sumio
  • Miura Noriyuki
  • Matsuda Kohei
  • Nagata Makoto
  • Bhasin Shivam
  • Mathieu Yves
  • Graba Tarik
  • Danger Jean-Luc
  IEEE Transactions on Computers, Institute of Electrical and Electronics Engineers, 2020, 69 (4), pp.534-548. (10.1109/TC.2019.2957355)
  DOI : 10.1109/TC.2019.2957355
• Compositional Languages Emerge in a Neural Iterated Learning Model
  
  • Ren Yi
  • Guo Shangmin
  • Labeau Matthieu
  • Cohen Shay B
  • Kirby Simon
  , 2020. The principle of compositionality, which enables natural language to represent complex concepts via a structured combination of simpler ones, allows us to convey an open-ended set of messages using a limited vocabulary. If compositionality is indeed a natural property of language, we may expect it to appear in communication protocols that are created by neural agents via grounded language learning. Inspired by the iterated learning framework, which simulates the process of language evolution, we propose an effective neural iterated learning algorithm that, when applied to interacting neural agents, facilitates the emergence of a more structured type of language. Indeed, these languages provide specific advantages to neural agents during training, which translates as a larger posterior probability, which is then incrementally amplified via the iterated learning procedure. Our experiments confirm our analysis, and also demonstrate that the emerged languages largely improve the generalization of the neural agent communication.
• Processor Anchor to Increase the Robustness Against Fault Injection and Cyber Attacks
  
  • Danger Jean-Luc
  • Facon Adrien
  • Guilley Sylvain
  • Heydemann Karine
  • Kühne Ulrich
  • Si Merabet Abdelmalek
  • Timbert Michaël
  • Pecatte Baptiste
  , 2021, 12244, pp.254-274. One major advance in software security would be to use robust processors which could assist the code developer to thwart both cyber and physical attacks. This paper presents a hardware-based solution which increases the security by checking the integrity of executed code on any microcontroller. Unlike other Control Flow Integrity (CFI) protections, this solution does not require modifications of the CPU pipeline, but relies on monitoring the interface between the processor and its instruction cache. The integrity of the execution flow and the instruction sequences (called Basic Blocks) is checked by hardware with precomputed metadata. Another module is dedicated to speed up the access to these metadata. This paper shows the effectiveness of the solution as the impact is as much as 21% in average on the execution time at the price of using memory space to store metadata along with the code. (10.1007/978-3-030-68773-1_12)
  DOI : 10.1007/978-3-030-68773-1_12
• Comment limiter les biais des algorithmes ?
  
  • With Charles-Albert
  • Lehalle Driss
  • Lamrani Marie
  • Brière David
  • Bounie Winston
  • Bertail Patrice
  • Clémençon Stéphan
  • Waelbroeck Patrick
  • Bounie David
  Les cahiers Louis Bachelier, Institut Louis Bachelier, 2020, 36. Although decision-support algorithms are now part of everyday life and are used in all sectors, especially finance, they are still criticized for their lack of transparency and their biased results. Researchers have studied this issue and make recommendations to reverse the situation.
• Experimental Analysis of the Electromagnetic Instruction Skip Fault Model
  
  • Menu Alexandre
  • Dutertre Jean-Max
  • Potin Olivier
  • Rigaud Jean-Baptiste
  • Danger Jean-Luc
  , 2020. Microcontrollers storing valuable data or using security functions are vulnerable to fault injection attacks. Among the various types of faults, instruction skips induced at runtime proved to be effective against identification routines or encryption algorithms. Until recently, most research works assessed a fault model that consists in a single instruction skip, i.e. the ability to prevent one chosen instruction in a program from being executed. We question this fault model for EM fault injection on experimental basis and report the possibility to induce several consecutive instructions skips. (10.1109/DTIS48698.2020.9081261)
  DOI : 10.1109/DTIS48698.2020.9081261
• A Dichotomy for Homomorphism-Closed Queries on Probabilistic Graphs
  
  • Amarilli Antoine
  • Ceylan İsmail İlkan
  , 2020. We study the problem of probabilistic query evaluation (PQE) over probabilistic graphs, namely, tuple-independent probabilistic databases (TIDs) on signatures of arity two. Our focus is the class of queries that is closed under homomorphisms, or equivalently, the infinite unions of conjunctive queries, denoted UCQ^\infty. Our main result states that all unbounded queries in UCQ^\infty are #P-hard for PQE. As bounded queries in UCQ^\infty are already classified by the dichotomy of Dalvi and Suciu [17], our results and theirs imply a complete dichotomy on PQE for UCQ^\infty queries over probabilistic graphs. This dichotomy covers in particular all fragments in UCQ^\infty such as negation-free (disjunctive) Datalog, regular path queries, and a large class of ontology-mediated queries on arity-two signatures. Our result is shown by reducing from counting the valuations of positive partitioned 2-DNF formulae (#PP2DNF) for some queries, or from the source-to-target reliability problem in an undirected graph (#U-ST-CON) for other queries, depending on properties of minimal models. (10.4230/LIPIcs.ICDT.2020.5)
  DOI : 10.4230/LIPIcs.ICDT.2020.5
• REMI: Mining Intuitive Referring Expressions on Knowledge Bases
  
  • Galárraga Luis
  • Delaunay Julien
  • Dessalles Jean-Louis
  , 2020, pp.387-390. A referring expression (RE) is a description that identifies a set of instances unambiguously. Mining REs from data finds applications in natural language generation, algorithmic journalism, and data maintenance. Since there may exist multiple REs for a given set of entities, it is common to focus on the most concise and informative (i.e., intuitive) ones. We present REMI, a method to mine intuitive REs on large knowledge bases. Our experimental evaluation shows that REMI finds REs deemed intuitive by users. Moreover we show that REMI is several orders of magnitude faster than an approach based on inductive logic programming. (10.5441/002/edbt.2020.39)
  DOI : 10.5441/002/edbt.2020.39
• Toward Visual Interactive Exploration of Heterogeneous Graphs
  
  • Burger Irène
  • Manolescu Ioana
  • Pietriga Emmanuel
  • Suchanek Fabian M
  , 2020. An interesting class of heterogeneous datasets, encountered for instance in data journalism applications, results from the inter-connection of data sources of different data models, ranging from very structured (e.g., relational or graphs) to semistructured (e.g., JSON, HTML, XML) to completely unstructured (text). Such heterogeneous graphs can be exploited e.g., by keyword search, to uncover connection between search keywords [1]. In this paper, we present a vision toward making such graphs easily comprehensible by human users, such as journalists seeking to understand and explore them. Our proposal is twofold: (i) abstracting the graph by recognizing structured entities; this simplifies the graph without information loss; (ii) relying on data visualization techniques to help users grasp the graph contents. Our work in this area continues; we present preliminary encouraging results.
• P-doping effect on external optical feedback dynamics in 1.3-microns InAs/GaAs quantum dot laser epitaxially grown on silicon
  
  • Dong Bozhang
  • Chen Jun-Da
  • Tsay Han-Ling
  • Huang Heming
  • Duan Jianan
  • Norman Justin C
  • Bowers John E
  • Lin Fan-Yi
  • Grillot Frédéric
  , 2020. This work reports on the optical feedback dynamics of InAs/GaAs QD lasers epitaxially grown on silicon operating in both the short and long delay regimes. Both undoped and p-doped QD lasers are considered. Whatever the external cavity length, no chaotic oscillations are observed on both samples as a result of the small α-factor observed in the silicon QD lasers. Despite that, experiments conducted in the short-cavity region raise period-one oscillation for the undoped QD laser. In addition, the transition from the short to long delay regimes can be finely covered by varying the external cavity length from 5 cm to 50 cm, and the boundaries associated to the appearance of the periodic oscillation are identified. In the short-cavity region, boundaries show some residual undulations resulting from interferences between internal and external cavity modes; whereas in the long-delay regime, the feedback ratio delimiting the boundaries keeps decreasing, until it progressively becomes rather independent of the external cavity length. Overall, our results showed that the p-doped device clearly exhibits a much higher tolerance to the different external feedback conditions than the undoped one, seeing that its periodic oscillation boundaries are barely impossible to retrieve at the maximum feedback strength of-7 dB. These results show for the first time the p-modulation doping effect on the enhancement of feedback insensitivity in both short-and long-delay configurations, which is of paramount importance for the development of ultra-stable silicon transmitters for photonic technologies. (10.1117/12.2555471)
  DOI : 10.1117/12.2555471