Publications

2020

• Excitability in Mid-Infrared Quantum Cascade Lasers: from Communication Jamming to Neuromorphic Photonics
  
  • Spitz O
  • Wu J
  • Carras M
  • Maisons G
  • Wong C W
  • Grillot F
  , 2020. We experimentally build a basic optical neuron by taking advantage of excitability in a semiconductor laser under optical feedback, rather than conventional injection schemes. This optical neuron operates faster than its biological and electronical counterparts.
• Experimental Demonstration of a 4D PDL-resilient Signaling for Long-haul Networks
  
  • Dumenil Arnaud
  • Awwad Elie
  • Measson Cyril
  • Le Gac Dylan
  , 2020.
• AADL: A Language to Specify the Architecture of Cyber-Physical Systems
  
  • Blouin Dominique
  • Borde Etienne
  , 2020.
• AUDIO-BASED AUTO-TAGGING WITH CONTEXTUAL TAGS FOR MUSIC
  
  • Ibrahim Karim M
  • Royo-Letelier Jimena
  • Epure Elena V.
  • Peeters Geoffroy
  • Richard Gael
  , 2020. Music listening context such as location or activity has been shown to greatly influence the users' musical tastes. In this work, we study the relationship between user context and audio content in order to enable context-aware music recommendation agnostic to user data. For that, we propose a semi-automatic procedure to collect track sets which leverages playlist titles as a proxy for context labelling. Using this, we create and release a dataset of ∼50k tracks labelled with 15 different contexts. Then, we present benchmark classification results on the created dataset using an audio auto-tagging model. As the training and evaluation of these models are impacted by missing negative labels due to incomplete annotations, we propose a sample-level weighted cross entropy loss to account for the confidence in missing labels and show improved context prediction results. (10.5281/zenodo.3648287)
  DOI : 10.5281/zenodo.3648287
• Approximate Bayesian computation with the sliced-Wasserstein distance
  
  • Nadjahi Kimia
  • de Bortoli Valentin
  • Durmus Alain
  • Badeau Roland
  • Şimşekli Umut
  , 2020. (10.1109/icassp40776.2020.9054735)
  DOI : 10.1109/icassp40776.2020.9054735
• Speech Intelligibility Enhancement by Equalization for in-Car Applications
  
  • Gentet Enguerrand
  • Bertrand David
  • Denjean Sebastien
  • Richard Gael
  • Roussarie Vincent
  , 2020, pp.6934-6938. (10.1109/ICASSP40776.2020.9053537)
  DOI : 10.1109/ICASSP40776.2020.9053537
• LEARNING TO RANK MUSIC TRACKS USING TRIPLET LOSS
  
  • Prétet Laure
  • Richard Gael
  • Peeters Geoffroy
  , 2020. Most music streaming services rely on automatic recommendation algorithms to exploit their large music catalogs. These algorithms aim at retrieving a ranked list of music tracks based on their similarity with a target music track. In this work, we propose a method for direct recommendation based on the audio content without explicitly tagging the music tracks. To that aim, we propose several strategies to perform triplet mining from ranked lists. We train a Convolutional Neural Network to learn the similarity via triplet loss. These different strategies are compared and validated on a large-scale experiment against an auto-tagging based approach. The results obtained highlight the efficiency of our system, especially when associated with an Auto-pooling layer.
• Neutral to Lombard Speech Conversion with Deep Learning
  
  • David Bertrand
  • Gentet Enguerrand
  • Denjean Sebastien
  • Richard Gael
  • Roussarie Vincent
  , 2020, pp.7739-7743. (10.1109/ICASSP40776.2020.9053006)
  DOI : 10.1109/ICASSP40776.2020.9053006
• DNN-Based Distributed Multichannel Mask Estimation for Speech Enhancement in Microphone Arrays
  
  • Furnon Nicolas
  • Serizel Romain
  • Illina Irina
  • Essid Slim
  , 2020. Multichannel processing is widely used for speech enhancement but several limitations appear when trying to deploy these solutions to the real-world. Distributed sensor arrays that consider several devices with a few microphones is a viable alternative that allows for exploiting the multiple devices equipped with microphones that we are using in our everyday life. In this context, we propose to extend the distributed adaptive node-specific signal estimation approach to a neural networks framework. At each node, a local filtering is performed to send one signal to the other nodes where a mask is estimated by a neural network in order to compute a global multi-channel Wiener filter. In an array of two nodes, we show that this additional signal can be efficiently taken into account to predict the masks and leads to better speech enhancement performances than when the mask estimation relies only on the local signals.
• Joint phoneme alignment and text-informed speech separation on highly corrupted speech
  
  • Schulze-Forster Kilian
  • Doire Clément
  • Richard Gael
  • Badeau Roland
  , 2020. Speech separation quality can be improved by exploiting textual information. However, this usually requires text-to-speech alignment at phoneme level. Classical alignment methods are made for rather clean speech and do not work as well on corrupted speech. We propose to perform text-informed speech-music separation and phoneme alignment jointly using recurrent neural networks and the attention mechanism. We show that it leads to benefits for both tasks. In experiments, phoneme transcripts are used to improve the perceived quality of separated speech over a non-informed baseline. Moreover, our novel phoneme alignment method based on the attention mechanism achieves state-of-the-art alignment accuracy on clean and on heavily corrupted speech.
• Audio-Based Detection of Explicit Content in Music
  
  • Vaglio Andrea
  • Hennequin Romain
  • Moussallam Manuel
  • Richard Gael
  • d'Alché-Buc Florence
  , 2020, pp.526-530. (10.1109/ICASSP40776.2020.9054278)
  DOI : 10.1109/ICASSP40776.2020.9054278
• Probabilistic filter and smoother for variational inference of Bayesian linear dynamical systems
  
  • Neri Julian
  • Badeau Roland
  • Depalle Philippe
  , 2020.
• How Confident Are You? Exploring The Role Of Fillers In The Automatic Prediction Of A Speaker’s Confidence
  
  • Dinkar Tanvi
  • Vasilescu Ioana
  • Pelachaud Catherine
  • Clavel Chloé
  , 2020, pp.8104-8108. Fillers", example "um" in English, have been linked to the "Feeling of Another's Knowing (FOAK)" or the listener's perception of a speaker's expressed confidence. Yet, in Spoken Language Processing (SLP) they remain unexplored, or overlooked as noise. We introduce a new and challenging task, that is the prediction of FOAK, which we think has widespread applicability, given the increasing popularity of automatic processing of educational and job interviews, reviews and speeches. We design a set of filler features based on linguistic literature, and investigate their potential in FOAK prediction. We show that the integration of information related to implicature meanings allows an improvement in the FOAK model and that the different functions of fillers are differently correlated with confidence. (10.1109/ICASSP40776.2020.9054374)
  DOI : 10.1109/ICASSP40776.2020.9054374
• A Prototypical Triplet Loss for Cover Detection
  
  • Doras Guillaume
  • Peeters Geoffroy
  , 2020, pp.3797-3801. Automatic cover detection - the task of finding in an audio dataset all covers of a query track - has long been a challenging theoretical problem in MIR community. It also became a practical need for music composers societies requiring to detect automatically if an audio excerpt embeds musical content belonging to their catalog. In a recent work, we addressed this problem with a convolutional neural network mapping each track's dominant melody to an embedding vector, and trained to minimize cover pairs distance in the embeddings space, while maximizing it for non-covers. We showed in particular that training this model with enough works having five or more covers yields state-of-the-art results. This however does not reflect the realistic use case, where music catalogs typically contain works with zero or at most one or two covers. We thus introduce here a new test set incorporating these constraints, and propose two contributions to improve our model's accuracy under these stricter conditions: we replace dominant melody with multi-pitch representation as input data, and describe a novel prototypical triplet loss designed to improve covers clustering. We show that these changes improve results significantly for two concrete use cases, large dataset lookup and live songs identification. (10.1109/ICASSP40776.2020.9054619)
  DOI : 10.1109/ICASSP40776.2020.9054619
• Laplace state space filter with exact inference and moment matching
  
  • Neri Julian
  • Depalle Philippe
  • Badeau Roland
  , 2020.
• Crypto Terminal: A New Open Device For Securing Blockchain Wallets
  
  • Urien Pascal
  , 2020, pp.1-3. (10.1109/ICBC48266.2020.9169410)
  DOI : 10.1109/ICBC48266.2020.9169410
• Dynamic-TDD Interference Tractability Approaches and Performance Analysis in Macro-Cell and Small-Cell Deployments
  
  • Nasri Ridha
  • Rachad Jalal
  • Decreusefond Laurent
  , 2020. Meeting the continued growth in data traffic volume, Dynamic Time Division Duplex (D-TDD) has been introduced as a solution to deal with the uplink (UL) and downlink (DL) traffic asymmetry, mainly observed for dense heterogeneous network deployments, since it is based on instantaneous traffic estimation and provide more resource assignment flexibility. However, the use of this feature requires new interference mitigation schemes capable to handle two additional types of interference between cells in opposite transmission direction: DL to UL and UL to DL interference. The aim of this work is to provide a complete analytical approach to model inter-cell interference in macro-cells deployment with dense small-cells. We derive the explicit expressions of Interference to Signal Ratio (ISR) at each position of the network, in both DL and UL, to quantify the impact of each type of interference on the perceived performance. Also, we provide the explicit expressions of the coverage probability as a functions of different system parameters by covering different scenarios. Finally, through system level simulations, we analyze the feasibility of D-TDD and its comparison with the static-TDD configuration.
• Multi-Tone Signal Optimization for Wireless Power Transfer in the Presence of Wireless Communication Links
  
  • Mouris Boules
  • Ghauch Hadi
  • Thobaben Ragnar
  • Jonsson B.
  IEEE Transactions on Wireless Communications, Institute of Electrical and Electronics Engineers, 2020, 19 (5), pp.3575-3590. (10.1109/TWC.2020.2974950)
  DOI : 10.1109/TWC.2020.2974950
• Sampling informative patterns from large single networks
  
  • Chehreghani Mostafa Haghir
  • Abdessalem Talel
  • Bifet Albert
  • Bouzbila Meriem
  Future Generation Computer Systems, Elsevier, 2020, 106, pp.653--658. The set of all frequent patterns that are extracted from a single network can be huge. A technique recently proposed for obtaining a compact, informative and useful set of patterns is output sampling, where a small set of frequent patterns is randomly chosen. However, existing output sampling algorithms work only in the transactional setting, where the database consists of a collection of relatively small graphs. In this paper, first we extend the output sampling framework to the single network setting where the database is a large single graph, counting supports of patterns is more complicated, and frequent patterns might be sampled based on any arbitrary target distribution. Then, we propose sampling techniques that are based on more interesting/informative measures or those that are specific to large single networks, such as product of the pattern size with its support, network compressibility, and pattern density. Finally, we study the empirical behavior of our algorithm in a real-world case study. (10.1016/J.FUTURE.2020.01.042)
  DOI : 10.1016/J.FUTURE.2020.01.042
• Sensor-Aided EMF Exposure Assessments in an Urban Environment Using Artificial Neural Networks
  
  • Wang Shanshan
  • Wiart Joe
  International Journal of Environmental Research and Public Health, MDPI, 2020, 17 (9), pp.3052:1-3052:15. This paper studies the time and space mapping of the electromagnetic field (EMF) exposure induced by cellular base station antennas (BSA) using artificial neural networks (ANN). The reconstructed EMF exposure map (EEM) in urban environment is obtained by using data from EMF sensor networks, drive testing and information accessible in a public database, e.g., locations and orientations of BSA. The performance of EEM is compared with Exposure Reference Map (ERM) based on simulations, in which parametric path loss models are used to reflect the complexity of urban cities. Then, a new hybrid ANN, which has the advantage of sorting and utilizing inputs from simulations efficiently, is proposed. Using both hybrid ANN and conventional regression ANN, the EEM is reconstructed and compared to the ERM first by the reconstruction approach considering only EMF exposure assessed from sensor networks, where the required number of sensors towards good reconstruction is explored; then, a new reconstruction approach using the sensors information combined with EMF along few streets from drive testing. Both reconstruction approaches use simulations to mimic measurements. The influence of city architecture on EMF exposure reconstruction is analyzed and the addition of noise is considered to test the robustness of ANN as well. (10.3390/ijerph17093052)
  DOI : 10.3390/ijerph17093052
• Efficient Batch-Incremental Classification Using UMAP for Evolving Data Streams
  
  • Bahri Maroua
  • Pfahringer Bernhard
  • Bifet Albert
  • Maniu Silviu
  , 2020, 12080, pp.40-53. Learning from potentially infinite and high-dimensional data streams poses significant challenges in the classification task. For instance, k-Nearest Neighbors (kNN) is one of the most often used algorithms in the data stream mining area that proved to be very resource-intensive when dealing with high-dimensional spaces. Uniform Manifold Approximation and Projection (UMAP) is a novel manifold technique and one of the most promising dimension reduction and visualization techniques in the non-streaming setting because of its high performance in comparison with competitors. However, there is no version of UMAP that copes with the challenging context of streams. To overcome these restrictions, we propose a batch-incremental approach that pre-processes data streams using UMAP, by producing successive embeddings on a stream of disjoint batches in order to support an incremental kNN classification. Experiments conducted on publicly available synthetic and real-world datasets demonstrate the substantial gains that can be achieved with our proposal compared to state-of-the-art techniques. (10.1007/978-3-030-44584-3_4)
  DOI : 10.1007/978-3-030-44584-3_4
• Percolation-Based Detection of Anomalous Subgraphs in Complex Networks
  
  • Larroche Corentin
  • Mazel Johan
  • Clémençon Stéphan
  , 2020, 12080, pp.287-299. The ability to detect an unusual concentration of extreme observations in a connected region of a graph is fundamental in a number of use cases, ranging from traffic accident detection in road networks to intrusion detection in computer networks. This task is usually performed using scan statistics-based methods, which require explicitly finding the most anomalous subgraph and thus are computationally intensive. We propose a more scalable method in the case where the observations are assigned to the edges of a large-scale network. The rationale behind our work is that if an anomalous cluster exists in the graph, then the subgraph induced by the most individually anomalous edges should contain an unexpectedly large connected component. We therefore reformulate our problem as the detection of anomalous sample paths of a percolation process on the graph, and our contribution can be seen as a generalization of previous work on percolation-based cluster detection. We evaluate our method through extensive simulations. (10.1007/978-3-030-44584-3_23)
  DOI : 10.1007/978-3-030-44584-3_23
• Benchmarks for Grid Flexibility Prediction: Enabling Progress and Machine Learning Applications
  
  • Kiedanski Diego
  • Kuntz Lauren
  • Kofman Daniel
  , 2020. Decarbonizing the grid is recognized worldwide as one of the objectives for the next decades. Its success depends on our ability to massively deploy renewable resources, but to fully benefit from those, grid flexibility is needed. In this paper we put forward the design of a benchmark that will allow for the systematic measurement of demand response programs' effectiveness, information that we do not currently have. Furthermore, we explain how the proposed benchmark will facilitate the use of Machine Learning techniques in grid flexibility applications.
• How Relevant is Hick's Law for HCI?
  
  • Liu Wanyu
  • Gori Julien
  • Rioul Olivier
  • Beaudouin-Lafon Michel
  • Guiard Yves
  , 2020, pp.1-11. Hick's law is a key quantitative law in Psychology that relates reaction time to the logarithm of the number of stimulus-response alternatives in a task. Its application to HCI is controversial: Some believe that the law does not apply to HCI tasks, others regard it as the cornerstone of interface design. The law, however, is often misunderstood. We review the choice-reaction time literature and argue that: (1) Hick's law speaks against, not for, the popular principle that 'less is better'; (2) logarithmic growth of observed temporal data is not necessarily interpretable in terms of Hick's law; (3) the stimulus-response paradigm is rarely relevant to HCI tasks, where choice-reaction time can often be assumed to be constant; and (4) for user interface design, a detailed examination of the effects on choice-reaction time of psychological processes such as visual search and decision making is more fruitful than a mere reference to Hick's law. (10.1145/3313831.3376878)
  DOI : 10.1145/3313831.3376878
• Welcome to the Course: Early Social Cues Influence Women's Persistence in Computer Science
  
  • Kizilcec Rene
  • Saltarelli Andrew
  • Bonfert-Taylor Petra
  • Goudzwaard Michael
  • Hamonic Ella
  • Sharrock Rémi
  , 2020, pp.1-13. First impressions influence subsequent behavior, especially when deciding how much effort to invest in an activity such as taking an online course. In computer programming courses, a context where social group stereotypes are salient, social cues early in the course can be used strategically to affirm members of historically underrepresented groups in their sense of belonging. We tested this idea in two randomized field experiments (N=53,922) by varying the social identity and status of the presenter of a welcome video and assessing online learners' persistence and achievement. Counter to our hypotheses, we found lower persistence among women in certain age groups if the welcome video was presented by a female instructor or by lower-status peers. Men remained unaffected. The results suggest that women are more responsive to social cues in online STEM courses, an environment where their social identity has been negatively stereotyped. Presenting a male and female instructor together was an effective strategy for retaining women in the course. (10.1145/3313831.3376752)
  DOI : 10.1145/3313831.3376752