DTSC - GTSA - Comunicaciones en congresos y otros eventos

Permanent URI for this collection

Browse

Recent Submissions

Now showing 1 - 20 of 32
  • Publication
    As light as your footsteps: design and evaluation of a portable device for changing body perception through a sound illusion
    (Sociedad Española De Acústica -Sea-, 2022-11-02) Prida Caballero, Daniel de la; Díaz Durán, Joaquín Roberto; Azpicueta Ruiz, Luis Antonio; Tajadura Jiménez, Ana; European Commission
    People’s body perception is highly malleable. Recent works have demonstrated that the dynamic modification of footstep sounds can lead people to perceive their body as thinner/lighter, walk more dynamically and feel happier, potentially supporting health. Previous studies modified the spectra of footstep sounds through a stereo 9-band analog graphic equalizer. While this system had minimal latency, it was not optimal as a wearable device, considering its weight (near 2 kg) and necessity of an electric outlet, which limited its applicability to real-world scenarios. Consequently, several substitute solutions were tested to improve portability, lightness and freedom of movement. For some, a non-satisfactory attempt was made to replicate the spectra of the original system. Therefore, it was hypothesized that a standalone digital microcomputer could increase portability and replicate the spectra. A novel device, using Bela.io and SuperCollider programming language, was tested, in which the spectral behavior of the original equalizer was replicated using cascaded biquad IIR filters. Objective and subjective experimental results suggest that, subject to the original system, we have successfully reduced weight and increased portability while keeping latency and spectral difference negligible. We foresee this novel system as a portable robust solution to induce illusory changes in body perception.
  • Publication
    Second-order asymptotics of Hoeffding-like hypothesis tests
    (IEEE, 2022-11-01) Kallumadatil Velluva, Harsha; Ravikumaran Nair, Jithin; Koch, Tobias Mirco; European Commission; Ministerio de Ciencia e Innovación (España)
    We consider a binary statistical hypothesis testing problem, where n independent and identically distributed random variables Z^n are either distributed according to the none hypothesis P or the alternate hypothesis Q, and only P is known. For this problem, a well-known test is the Hoeffding test, which accepts P if the Kullback-Leibler (KL) divergence between the empirical distribution of Z^n and P is below some threshold. In this paper, we consider Hoeffding-like tests, where the KL divergence is replaced by other divergences, and characterize, for a large class of divergences, the first and second-order terms of the type-II error for a fixed type-I error. Since the considered class includes the KL divergence, we obtain the second-order term of the Hoeffding test as a special case.
  • Publication
    Convolutional neural networks to identify drivers in persistent atrial fibrilation patients
    (Elsevier, 2022-05) Ríos Muñoz, Gonzalo Ricardo; Ávila Alonso, Pablo; Carta Bergaz, Alejandro; Soto Flores, Nina; González-Torrecilla, Esteban; Atienza, Felipe; Fernández Avilés, Francisco; Arenal Maíz, Ángel
    Background. Identifying driving and initiating areas in persistent atrial fibrillation (AF) might contribute to better understanding AF and improve the ablation treatment. However, their correct identification still requires visual inspection or heavy signal processing algorithms that sometimes distort the electrograms (EGMs) information. In this line, artificial intelligence has recently proven to be a powerful tool that outperforms current detection methods. Objective. To automatically detect AF rotational activity (RA) drivers (rotors) with a convolutional neural network that employs raw multi-electrode EGMs without signal pre-processing. Methods. We trained 2 different CNN-based models, using 44,660 unipolar and bipolar EGMs respectively. EGMs were acquired with a multi-electrode catheter from 49 persistent AF patients. RA was annotated by an automated algorithm based on the local activation times of unipolar EGMs acquired with a 20-pole catheter. This annotation involved time-demanding signal pre-processing and post-processing steps. The models implemented recurrent CNN-based layers with a sigmoid output layer for detecting RA or no-RA Results. The CNN model trained with bipolar EGMs exhibited better accuracy than the unipolar EGMs for the test data (80.04 vs 68.40 respectively). Precision results were similar, 74.14 vs 77.91, and bipolar recall was greater than the unipolar, 92.27 vs 51.36. Conclusion. The CNN-based model allows RA driver assessment in AF patients without computationally heavy and time-consuming algorithms based on traditional signal pre-processing methods. Bipolar EGMs exhibited the best performance even though the training features and labels came from unipolar EGM data. The model could be used in real-time in new catheter ablation strategies to identify atrial substrate driving AF faster than other methods.
  • Publication
    Modeling phone call durations via switching Poisson processes with applications in mental health
    (IEEE, 2020-10-20) Bonilla Escribano, Pablo; Ramírez García, David; Artés Rodríguez, Antonio
    This work models phone call durations via switching Poisson point processes. This kind of processes is composed by two intertwined intensity functions: one models the start of a call, whereas the other one models when the call ends. Thus, the call duration is obtained from the inverse of the intensity function of finishing a call. Additionally, to model the circadian rhythm present in human behavior, we shall use a (pos-itive) truncated Fourier series as the parametric form of the intensities. Finally, the maximum likelihood estimates of the intensity functions are obtained using a trust region method and the performance is evaluated on synthetic and real data, showing good results.
  • Publication
    On joint detection and decoding in short-packet communications
    (IEEE, 2021-12-07) Lancho Serrano, Alejandro; Ostman, Johan; Giuseppe, Durisi
    We consider a communication problem in which the receiver must first detect the presence of an information packet and, if detected, decode the message carried within it. We present general nonasymptotic upper and lower bounds on the maximum coding rate that depend on the blocklength, the probability of false alarm, the probability of misdetection, and the packet er-ror probability. The bounds, which are expressed in terms of binary-hypothesis-testing performance metrics, generalize finite-blocklength bounds derived previously for the scenario when a genie informs the receiver whether a packet is present. The bounds apply to detection performed either jointly with decoding on the entire data packet, or separately on a dedicated preamble. The results presented in this paper can be used to determine the block-length values at which the performance of a communication system is limited by its ability to perform packet detection satisfacto-rily, and to assess the difference in performance between preamble-based detection, and joint detection and decoding. Numerical re-sults pertaining to the binary-input AWGN channel are provided.
  • Publication
    Hidden Markov Models for Activity Detection in Atrial Fibrillation Electrograms
    (IEEE, 2020-09-13) Ríos Muñoz, Gonzalo Ricardo; Moreno Pino, Fernando; Soto, Nina; Martínez Olmos, Pablo; Artés Rodríguez, Antonio; Fernández Avilés, Francisco; Arenal, Ángel; Comunidad de Madrid; Ministerio de Ciencia, Innovación y Universidades (España)
    Activity detection in atrial fibrillation (AF) electrograms (EGMs) is a key concept to understand the mechanisms of this frequent arrhythmia and design new strategies for its treatment. We present a new method that employs Hidden Markov Models (HMMs) to identify activity presence in bipolar EGMs. The method is fully unsupervised and hence it does not require labeled training data. The HMM activity detection method was validated and compared to the non-linear energy operator (NLEO) method for a set of manually annotated EGMs. The HMM performed better than the NLEO and exhibited more robustness in the presence of low voltage fragmented EGMs.
  • Publication
    Particle Filter Tracking of Complex Stochastic Systems Applied to In Silico Wavefront Propagation
    (IEEE, 2019-06-24) Ríos Muñoz, Gonzalo Ricardo; Artés Rodríguez, Antonio; Míguez Arenas, Joaquín; Comunidad de Madrid; Ministerio de Economía y Competitividad (España)
    A high dimensional tracking system based on the FithzHugh-Nagumo (FH-N) equations emulating the biological excitation and propagation dynamics of the action potential across cardiac cells is proposed. The modified FH-N model tracks the electric cardiac wavefronts on a tissue, emulating an approximated atrial fibrillation scenario. Bayesian tracking is achieved with two particle filter (PF) schemes: a sequential Auxiliary PF (APF) and a parallelized method, Independent APF (IAPF). The numerical results of the two examples, involving both estimation errors and running times, provide numerical evidence that support the theoretical findings.
  • Publication
    Patient-Tailored In Silico 3D Simulations and Models From Electroanatomical Maps of the Left Atrium
    (IEEE, 2019-06-24) Ríos Muñoz, Gonzalo Ricardo; Rocher, Sara; Artés Rodríguez, Antonio; Arenal, Ángel; Saiz, Javier; Sánchez, Carlos; Comunidad de Madrid; Ministerio de Economía y Competitividad (España)
    The mechanisms underlying atrial fibrillation (AF) are still under debate, making treatments for this arrhythmia remain suboptimal, with most treatments applied in a standard fashion with no patient personalization. Recent technological advances in electroanatomical mapping (EAM) using multi-electrode catheter allow the physicians to better characterize the substrate, thanks to a better spatial resolution and higher density of acquisition points. Taking advantage of this technology, we describe a workflow to build personalized electrophysiological atrial models for AF patients. We seek to better predict the outcome of a treatment and study the AF problem in a more specific scenario. We generated physiological 3D models from the EAM data using hexahedral meshing of element size 300μm, and added fiber orientation based on a generic model. We used the local activation time (LAT) maps performed in sinus rhythm (SR) to estimate the conduction velocity (CV) of the regions in the atrium with a new method that combines the LATs of neighboring tissue as the average CV of triplets of points. We also characterized the cellular model by Maleckar et al. in terms of longitudinal conductivity and CV to personalize the atrial models. We were able to simulate SR and AF scenarios on the personalized models, and we generated a database of atrial models for future analysis.
  • Publication
    Causality analysis of atrial fibrillation electrograms
    (IEEE, 2015-09-06) Luengo García, David; Ríos Muñoz, Gonzalo Ricardo; Elvira Arregui, Víctor; Ministerio de Economía y Competitividad (España)
    Multi-channel intracardiac electrocardiograms (electrograms) are sequentially acquired during heart surgery performed on patients with sustained atrial fibrillation (AF) to guide radio frequency catheter ablation. These electrograms are used by cardiologists to determine candidate areas for ablation (e.g., areas corresponding to high dominant frequencies or complex electrograms). In this paper, we introduce a novel hierarchical causality analysis method for the multi-output sequentially acquired electrograms. The causal model obtained provides important information regarding delays among signals as well as the direction and strength of their causal connections. The tool developed may ultimately serve to guide cardiologists towards candidate areas for catheter ablation. Preliminary results on synthetic signals are used to validate the proposed approach.
  • Publication
    Scaling laws for many-access channels and unsourced random access
    (IEEE, 2022-03-04) Ravikumaran Nair, Jithin; Koch, Tobias Mirco; European Commission; Ministerio de Ciencia e Innovación (España)
    In the emerging Internet of Things, a massive number of devices may connect to one common receiver. Consequently, models that study this setting are variants of the classical multiple-access channel where the number of users grows with the blocklength. Roughly, these models can be classified into three groups based on two criteria: the notion of probability of error and whether users use the same codebook. The first group follows the classical notion of probability of error and assumes that users use different codebooks. In the second group, users use different codebooks, but a new notion of probability of error called per-user probability of error is considered. The third group considers the per-user probability of error and that users are restricted to use the same codebook. This group is also known as unsourced random access. For the first and second groups of models, scaling laws that describe the capacity per unit-energy as a function of the order of growth of users were characterized by Ravi and Koch (arxiv.org/abs/2012.10350). In this paper, we first review these results. We then present scaling laws for the third group of models, i.e., unsourced random access.
  • Publication
    Demand-Private Coded Caching and the Exact Trade-off for N=K=2
    (IEEE, 2020-02-21) Kamath, Sneha; Ravikumaran Nair, Jithin; Dey, Bikash Kumar; European Commission
    The distributed coded caching problem has been studied extensively in the recent past. While the known coded caching schemes achieve an improved transmission rate, they violate the privacy of the users since in these schemes the demand of one user is revealed to others in the delivery phase. In this paper, we consider the coded caching problem under the constraint that the demands of the other users remain information theoretically secret from each user. We first show that the memory-rate pair (M, min{ N, K}(1- M / N)) is achievable under information theoretic demand privacy, while using broadcast transmissions. Using this, we show that perfectly demand-private coded caching rate is order optimal for all parameter regimes. We then show that a demand-private scheme for N files and K users can be obtained from a non-private scheme that satisfies only a restricted subset of demands of N K users for N files. We then focus on the demand-private coded caching problem for K = 2 users, N = 2 files. We characterize the exact memory-rate trade-off for this case. To show the achievability, we use our first result to construct a demand-private scheme from a non-private scheme satisfying a restricted demand subset that is known from an earlier work by Tian. Further, by giving a converse based on the extra requirement of privacy, we show that the obtained achievable region is the exact memory-rate trade-off.
  • Publication
    Improved memory-rate trade-off for caching with demand privacy
    (IEEE, 2021-04-11) Gurjarpadhye, Chinmay; Ravikumaran Nair, Jithin; Dey, Bikash Kumar; Karamchandani, Nikhil; European Commission
    We consider the demand-private coded caching problem in a noiseless broadcast network. It is known from past works that a demand-private scheme for N files and K users can be obtained from a non-private scheme for N files and NK users. We first propose a scheme that improves on this idea by removing some redundant transmissions. The memory- rate trade-off achieved using this scheme is shown to be within a multiplicative factor of 3 from the optimal for all the memory regimes when K K = 2.
  • Publication
    On the error probability of optimal codes in Gaussian channels under maximal power constraint
    (IEEE, 2019-07-07) Vázquez Vilar, Gonzalo; European Commission; Ministerio de Economía y Competitividad (España)
    For an additive white Gaussian noise channel, we prove that Th. 41 in [Polyanskiy, Poor, Verdά2010] is a lower bound to the error probability of any channel code satisfying the maximal power constraint. In contrast, the (tighter) lower bound to the error probability in Eq. (20) in [Shannon 1959] only holds under equal power constraint.
  • Publication
    Improving deep learning performance with missing values via deletion and compensation
    (Springer Nature, 2020-09) Sánchez Morales, Adrián; Sancho Gomez, Jose Luis; Martinez Garcia, Juan Antonio; Figueiras, Aníbal; Ministerio de Economía y Competitividad (España)
    Missing values in a dataset is one of the most common difficulties in real applications. Many different techniques based on machine learning have been proposed in the literature to face this problem. In this work, the great representation capability of the stacked denoising auto-encoders is used to obtain a new method of imputating missing values based on two ideas: deletion and compensation. This method improves imputation performance by artificially deleting values in the input features and using them as targets in the training process. Nevertheless, although the deletion of samples is demonstrated to be really efficient, it may cause an imbalance between the distributions of the training and the test sets. In order to solve this issue, a compensation mechanism is proposed based on a slight modification of the error function to be optimized. Experiments over several datasets show that the deletion and compensation not only involve improvements in imputation but also in classification in comparison with other classical techniques.
  • Publication
    Generalized CMAC adaptive ensembles for concept-drifting data streams
    (IEEE, 2017-08-28) González Serrano, Francisco Javier; Figueiras, Aníbal; Comunidad de Madrid; Ministerio de Economía y Competitividad (España)
    In this paper we propose to use an adaptive ensemble learning framework with different levels of diversity to handle streams of data in non-stationary scenarios in which concept drifts are present. Our adaptive system consists of two ensembles, each one with a different level of diversity (from high to low), and, therefore, with different and complementary capabilities, that are adaptively combined to obtain an overall system of improved performance. In our approach, the ensemble members are generalized CMACs, a linear-in-the-parameters network. The ensemble of CMACs provides a reasonable trade-off between expressive power, simplicity, and fast learning speed. At the end of the paper, we provide a performance analysis of the proposed learning framework on benchmark datasets with concept drifts of different levels of severity and speed.
  • Publication
    A High-SNR Normal Approximation for MIMO Rayleigh Block-Fading Channels
    (IEEE, 2020-08-24) Qi, Chao; Koch, Tobias Mirco; European Commission; Ministerio de Economía y Competitividad (España)
    This paper concerns the maximum coding rate at which a code of given blocklength can be transmitted with a given block-error probability over a non-coherent Rayleigh block-fading channel with multiple transmit and receive antennas (MIMO). In particular, a high-SNR normal approximation of the maximum coding rate is presented, which is proved to become accurate as the signal-to-noise ratio (SNR) and the number of coherence intervals L tend to infinity.
  • Publication
    Bursty Wireless Networks of Bounded Capacity
    (IEEE, 2020-08-24) Villacrés Estrada, Grace Silvana; Koch, Tobias Mirco; Vázquez Vilar, Gonzalo; European Commission; Ministerio de Economía y Competitividad (España)
    The channel capacity of wireless networks Is often studied under the assumption that the communicating nodes have perfect channel-state information and that interference is always present. In this paper, we study the channel capacity of a wireless network without these assumptions, i.e., a bursty noncoherent wireless network where the users are grouped in cells and the base-station features several receive antennas. We demonstrate that the channel capacity is bounded in the signal-to-noise ratio (SNR) when the number of receive antennas is finite and the probability of presence of interference is strictly positive.
  • Publication
    Capacity per Unit-Energy of Gaussian Random Many-Access Channels
    (IEEE, 2020-08-24) Ravikumaran Nair, Jithin; Koch, Tobias Mirco; European Commission; Ministerio de Economía y Competitividad (España)
    We consider a Gaussian multiple-access channel with random user activity where the total number of userslₙ and the average number of active users kₙ may be unbounded. For this channel, we characterize the maximum number of bits that can be transmitted reliably per unit-energy in terms of lₙ and kₙ . We show that if kₙ log lₙ is sublinear in n, then each user can achieve the single-user capacity per unit-energy. Conversely, if kₙ log lₙ is superlinear in n, then the capacity per unit-energy is zero. We further demonstrate that orthogonal-access schemes, which are optimal when all users are active with probability one, can be strictly suboptimal.
  • Publication
    Finite-Blocklength Approximations for Noncoherent Rayleigh Block-Fading Channels
    (IEEE, 2020-03-30) Lancho Serrano, Alejandro; Östman, Johan; Koch, Tobias Mirco; Vázquez Vilar, Gonzalo; European Commission; Ministerio de Economía y Competitividad (España)
    Several emerging wireless communication services and applications have stringent latency requirements, necessitating the transmission of short packets. To obtain performance benchmarks for short-packet wireless communications, it is crucial to study the maximum coding rate as a function of the blocklength, commonly called finite-blocklength analysis. A finiteblocklength analysis can be performed via nonasymptotic bounds or via refined asymptotic approximations. This paper reviews finite-blocklength approximations for the noncoherent Rayleigh block-fading channel. These approximations have negligible computational cost compared to the nonasymptotic bounds and are shown to be accurate for error probabilities as small as 10-8 [super index] and SNRs down to 0 dB.
  • Publication
    Capacity per Unit-Energy of Gaussian Many-Access Channels
    (IEEE, 2019-07-07) Ravikumaran Nair, Jithin; Koch, Tobias Mirco; European Commission; Ministerio de Economía y Competitividad (España)
    We consider a Gaussian multiple-access channel where the number of transmitters grows with the blocklength n. For this setup, the maximum number of bits that can be transmitted reliably per unit-energy is analyzed. We show that if the number of users is of an order strictly above n/log n, then the users cannot achieve any positive rate per unit-energy. In contrast, if the number of users is of order strictly below n/log n, then each user can achieve the single-user capacity per unit-energy (log e)/N 0 (where N 0 /2 is the noise power) by using an orthogonal access scheme such as time division multiple access. We further demonstrate that orthogonal codebooks, which achieve the capacity per unit-energy when the number of users is bounded, can be strictly suboptimal.