Search

We present a comprehensive study of the star formation histories of massive-quenched galaxies at $z=3$ in 3 semi-analytic models (SHARK, GAEA, GALFORM) and 3 cosmological hydrodynamical simulations (EAGLE, Illustris-TNG, Simba). We study the predicted number density and stellar mass function of massive-quenched galaxies, their formation and quenching timescales and star-formation properties of their progenitors. Predictions are disparate in all these diagnostics, for instance: (i) some simulations reproduce the observed number density of very massive-quenched galaxies ($>10^{11}\rm M_{\odot}$) but underpredict the high density of intermediate-mass ones, while others fit well the lower masses but underpredict the higher ones; (ii) In most simulations, except for GAEA and EAGLE, most massive-quenched galaxies had starburst periods, with the most intense ones happening at $4300\rm M_{\odot}\,yr^{-1}$; (iii) quenching timescales are in the range $\approx 20-150$~Myr depending on the simulation; among other differences. These disparate predictions can be tied to the adopted Active Galactic Nuclei (AGN) feedback model. For instance, the explicit black-hole (BH) mass dependence to trigger the "radio mode" in Illustris-TNG and Simba makes it difficult to produce quenched galaxies with intermediate stellar masses, also leading to higher baryon collapse efficiencies ($\approx 15-30$%); while the strong bolometric luminosity dependence of the AGN outflow rate in GAEA leads to BHs of modest mass quenching galaxies. Current observations are unable to distinguish between these different predictions due to the small sample sizes. However, these predictions are testable with current facilities and upcoming observations, allowing a "true physics experiment" to be carried out., Comment: Submitted for publication in MNRAS. 32 pages (21 of main body and 11 of appendices). Comments welcome!

This paper presents the work carried out by the ASASVIcomtech team, made up of researchers from Vicomtech and University of Granada, for the ASVspoof5 Challenge. The team has participated in both Track 1 (speech deepfake detection) and Track 2 (spoofing-aware speaker verification). This work started with an analysis of the challenge available data, which was regarded as an essential step to avoid later potential biases of the trained models, and whose main conclusions are presented here. With respect to the proposed approaches, a closed-condition system employing a deep complex convolutional recurrent architecture was developed for Track 1, although, unfortunately, no noteworthy results were achieved. On the other hand, different possibilities of open-condition systems, based on leveraging self-supervised models, augmented training data from previous challenges, and novel vocoders, were explored for both tracks, finally achieving very competitive results with an ensemble system., Comment: This paper was accepted at ASVspoof Workshop 2024

We propose and demonstrate a compact and simple vector bending sensor capable of distinguishing any direction and amplitude with high accuracy. The sensor consists of a short segment of asymmetric multicore fiber (MCF) fusion spliced to a standard single mode fiber. The reflection spectrum of such a structure shifts and shrinks in specific manners depending on the direction in which the MCF is bent. By monitoring simultaneously wavelength shift and light power variations, the amplitude and bend direction of the MCF can be unmistakably measured in any orientation, from 0{\deg} to 360{\deg}. The bending sensor proposed here is highly sensitive even for small bending angles (below 1{\deg}).

In this paper, we report on a multicore fiber-based (MCF) temperature sensor that operates in a wide thermal range and that is robustly packaged to withstand harsh environments. To develop the sensor, the fundamentals concerning the effect of temperature on such fibers have been analyzed in detail to predict the most temperature sensitive MCF geometry. Thanks to it, the device, which operates in reflection mode and consists of a short segment of strongly coupled MCF fusion spliced to a standard single mode fiber, shows higher sensitivity than other devices with identical configuration. Regarding its packaging, it consists of an inner ceramic and two outer metallic tubes to provide rigidity and protection against impacts or dirt. The device was calibrated for a thermal range from 25 C to 900 C and a K-type thermocouple was used as reference. Our results suggest that the manufactured optical thermometer is as accurate as the electronic one, reaching a sensitivity up to 29.426 pm/C with the advantage of being passive, compact and easy to fabricate and interrogate. Therefore, we believe this device is appealing for industrial applications that require highly sensitive temperature sensing in very demanding environments, and that the analysis included in this work could be analogously applied to develop sensitivity-optimized devices for other parameters of interest.

The present work deals with a curvature sensor that consists of two segments of asymmetric multicore fiber (MCF) fusion spliced with standard single mode fiber (SMF). The MCF comprises three strongly coupled cores; one of such cores is at the geometrical center of the MCF. The two segments of MCF are short, have different lengths (less than 2 cm each), and are rotated 180 degrees with respect to each other. The fabrication of the sensor was carried out with a fusion splicing machine that has the means for rotating optical fibers. It is demonstrated that the sensor behaves as two SMF-MCF-SMF structures in series, and consequently, it has enhanced sensitivity. The device proposed here can be used to sense the direction and amplitude of curvature by monitoring either wavelength shifts or intensity changes. In the latter case, high curvature sensitivity was observed. The device can also be used for the development of other highly sensitive sensors to monitor, for example, vibrations, force, pressure, or any other parameter that induces periodic or local curvature or bending to the MCF segments.

We report on a compact, highly sensitive all-fiber accelerometer suitable for low frequency and low amplitude vibration sensing. The sensing elements in the device are two short segments of strongly coupled asymmetric multicore fiber (MCF) fusion spliced at 180{\deg} with respect to each other. Such segments of MCF are sandwiched between standard single mode fibers. The reflection spectrum of the device exhibits a narrow spectrum whose height and position in wavelength changes when it is subjected to vibrations. The interrogation of the accelerometer was carried out by a spectrometer and a photodetector to measure simultaneously wavelength shift and light power variations. The device was subjected to a wide range of vibration frequencies, from 1 mHz to 30 Hz, and accelerations from 0.76 mg to 29.64 mg, and performed linearly, with a sensitivity of 2.213 nW/mg. Therefore, we believe the accelerometer reported here may represent an alternative to existing electronic and optical accelerometers, especially for low frequency and amplitude vibrations, thanks to its compactness, simplicity, cost-effectiveness, implementation easiness and high sensitivity.

In this work, we report on a twin-core fiber sensor system that provides improved spectral efficiency, allows for multiplexing and gives low level of crosstalk. Pieces of the referred strongly coupled multicore fiber are used as sensors in a laser cavity incorporating a pulsed semiconductor optical amplifier (SOA). Each sensor has its unique cavity length and can be addressed individually by electrically matching the periodic gating of the SOA to the sensors cavity roundtrip time. The interrogator acts as a laser and provides a narrow spectrum with high signal-to-noise ratio. Furthermore, it allows distinguishing the response of individual sensors even in the case of overlapping spectra. Potentially, the number of interrogated sensors can be increased significantly, which is an appealing feature for multipoint sensing.

Coherent perfect absorption (CPA) is an interference process associated with the zeros of the scattering matrix that enables light-with-light interactions in linear systems, of interest for optical computing, data processing and sensing. However, the noise properties of CPA remain relatively unexplored. Here, we demonstrate that CPA thermal noise signals exhibit a unique property: they are orthogonal to the signals transmitted through the network. In turn, such property enables a variety of thermal noise management effects, such as the physical separability of thermal noise and transmitted signals, and "externally lossless" networks that internally host radiative heat transfer processes. We believe that our results provide a new perspective on the many CPA technologies currently under development.

We introduce version 2.0 of the SHARK semi-analytic model of galaxy formation after many improvements to the physics included. The most significant being: (i) a model describing the exchange of angular momentum (AM) between the interstellar medium and stars; (ii) a new active galactic nuclei feedback model which has two modes, a wind and a jet mode, with the jet mode tied to the jet energy production; (iii) a model tracking the development of black hole (BH) spins; (iv) more sophisticated modelling of environmental effects on satellite galaxies; and (v) automatic parameter exploration using Particle Swarm Optimisation. We focus on two timely research topics: the structural properties of galaxies and the quenching of massive galaxies. For the former, SHARK v2.0 is capable of producing a more realistic stellar size-mass relation with a plateau marking the transition from disk- to bulge-dominated galaxies, and scaling relations between specific AM and mass that agree well with observations. For the quenching of massive galaxies, SHARK v2.0 produces massive galaxies that are more quenched than the previous version, reproducing well the observed relations between star formation rate (SFR) and stellar mass, and specific SFR and BH mass at $z=0$. SHARK v2.0 produces a number density of massive-quiescent galaxies >1dex higher than the previous version, in good agreement with JWST observations at $z\le 5$; predicts a stellar mass function of passive galaxies in reasonably good agreement with observations at $0.5

Flare frequency distributions represent a key approach to addressing one of the largest problems in solar and stellar physics: determining the mechanism that counter-intuitively heats coronae to temperatures that are orders of magnitude hotter than the corresponding photospheres. It is widely accepted that the magnetic field is responsible for the heating, but there are two competing mechanisms that could explain it: nanoflares or Alfv\'en waves. To date, neither can be directly observed. Nanoflares are, by definition, extremely small, but their aggregate energy release could represent a substantial heating mechanism, presuming they are sufficiently abundant. One way to test this presumption is via the flare frequency distribution, which describes how often flares of various energies occur. If the slope of the power law fitting the flare frequency distribution is above a critical threshold, $\alpha=2$ as established in prior literature, then there should be a sufficient abundance of nanoflares to explain coronal heating. We performed $>$600 case studies of solar flares, made possible by an unprecedented number of data analysts via three semesters of an undergraduate physics laboratory course. This allowed us to include two crucial, but nontrivial, analysis methods: pre-flare baseline subtraction and computation of the flare energy, which requires determining flare start and stop times. We aggregated the results of these analyses into a statistical study to determine that $\alpha = 1.63 \pm 0.03$. This is below the critical threshold, suggesting that Alfv\'en waves are an important driver of coronal heating., Comment: 1,002 authors, 14 pages, 4 figures, 3 tables, published by The Astrophysical Journal on 2023-05-09, volume 948, page 71

The formation of galaxies by gradual hierarchical co-assembly of baryons and cold dark matter halos is a fundamental paradigm underpinning modern astrophysics and predicts a strong decline in the number of massive galaxies at early cosmic times. Extremely massive quiescent galaxies (stellar masses $>10^{11}$ M$_\odot$) have now been observed as early as 1-2 billions years after the Big Bang; these are extremely constraining on theoretical models as they form 300-500 Myr earlier and only some models can form massive galaxies this early. Here we report on the spectroscopic observations with the James Webb Space Telescope of a massive quiescent galaxy ZF-UDS-7329 at redshift 3.205 $\pm$ 0.005 that eluded deep ground-based spectrscopy, is significantly redder than typical and whose spectrum reveals features typical of much older stellar populations. Detailed modelling shows the stellar population formed around 1.5 billion years earlier in time (z ~ 11) at an epoch when dark matter halos of sufficient hosting mass have not yet assembled in the standard scenario. This observation may point to the presence of undetected populations of early galaxies and the possibility of significant gaps in our understanding of early stellar populations, galaxy formation and/or the nature of dark matter., Comment: 21 pages, 6 figures, updated to reflect accepted version. v3 with figure caption clarified

Industry 4.0 has brought numerous advantages, such as increasing productivity through automation. However, it also presents major cybersecurity issues such as cyberattacks affecting industrial processes. Federated Learning (FL) combined with time-series analysis is a promising cyberattack detection mechanism proposed in the literature. However, the fact of having a single point of failure and network bottleneck are critical challenges that need to be tackled. Thus, this article explores the benefits of the Decentralized Federated Learning (DFL) in terms of cyberattack detection and resource consumption. The work presents TemporalFED, a software module for detecting anomalies in industrial environments using FL paradigms and time series. TemporalFED incorporates three components: Time Series Conversion, Feature Engineering, and Time Series Stationary Conversion. To evaluate TemporalFED, it was deployed on Fedstellar, a DFL framework. Then, a pool of experiments measured the detection performance and resource consumption in a chemical gas industrial environment with different time-series configurations, FL paradigms, and topologies. The results showcase the superiority of the configuration utilizing DFL and Semi-Decentralized Federated Learning (SDFL) paradigms, along with a fully connected topology, which achieved the best performance in anomaly detection. Regarding resource consumption, the configuration without feature engineering employed less bandwidth, CPU, and RAM than other configurations.

In 2016, Google proposed Federated Learning (FL) as a novel paradigm to train Machine Learning (ML) models across the participants of a federation while preserving data privacy. Since its birth, Centralized FL (CFL) has been the most used approach, where a central entity aggregates participants' models to create a global one. However, CFL presents limitations such as communication bottlenecks, single point of failure, and reliance on a central server. Decentralized Federated Learning (DFL) addresses these issues by enabling decentralized model aggregation and minimizing dependency on a central entity. Despite these advances, current platforms training DFL models struggle with key issues such as managing heterogeneous federation network topologies. To overcome these challenges, this paper presents Fedstellar, a platform extended from p2pfl library and designed to train FL models in a decentralized, semi-decentralized, and centralized fashion across diverse federations of physical or virtualized devices. The Fedstellar implementation encompasses a web application with an interactive graphical interface, a controller for deploying federations of nodes using physical or virtual devices, and a core deployed on each device which provides the logic needed to train, aggregate, and communicate in the network. The effectiveness of the platform has been demonstrated in two scenarios: a physical deployment involving single-board devices such as Raspberry Pis for detecting cyberattacks, and a virtualized deployment comparing various FL approaches in a controlled environment using MNIST and CIFAR-10 datasets. In both scenarios, Fedstellar demonstrated consistent performance and adaptability, achieving F1 scores of 91%, 98%, and 91.2% using DFL for detecting cyberattacks and classifying MNIST and CIFAR-10, respectively, reducing training time by 32% compared to centralized approaches.

Time-varying media break temporal symmetries while preserving spatial symmetries intact. Thus, it represents an excellent conceptual framework to investigate the fundamental implications of Noether's theorem for the electromagnetic field. At the same time, addressing momentum conservation in time-varying media sheds light on the Abraham-Minkowski debate, where two opposing forms of the electromagnetic field momentum are defended. Here, we present a tutorial review on the conservation of momentum in time-varying media. We demonstrate that the Minkowski momentum is a conserved quantity with three independent approaches of increasing complexity: (i) via the application of the boundary conditions for Maxwell equations at a temporal boundary, (ii) testing for constants of motion and deriving conservation laws, and (iii) applying temporal and spatial translations within the framework of the Lagrangian theory of the electromagnetic field. Each approach provides a different and complementary insight into the problem.

The COVID-19 pandemic has led to the saturation of public health services worldwide. In this scenario, the early diagnosis of SARS-Cov-2 infections can help to stop or slow the spread of the virus and to manage the demand upon health services. This is especially important when resources are also being stretched by heightened demand linked to other seasonal diseases, such as the flu. In this context, the organisers of the DiCOVA 2021 challenge have collected a database with the aim of diagnosing COVID-19 through the use of coughing audio samples. This work presents the details of the automatic system for COVID-19 detection from cough recordings presented by team PANACEA. This team consists of researchers from two European academic institutions and one company: EURECOM (France), University of Granada (Spain), and Biometric Vox S.L. (Spain). We developed several systems based on established signal processing and machine learning methods. Our best system employs a Teager energy operator cepstral coefficients (TECCs) based frontend and Light gradient boosting machine (LightGBM) backend. The AUC obtained by this system on the test set is 76.31% which corresponds to a 10% improvement over the official baseline., Comment: Accepted in INTERSPEECH 2021

This review summarises the status of silent speech interface (SSI) research. SSIs rely on non-acoustic biosignals generated by the human body during speech production to enable communication whenever normal verbal communication is not possible or not desirable. In this review, we focus on the first case and present latest SSI research aimed at providing new alternative and augmentative communication methods for persons with severe speech disorders. SSIs can employ a variety of biosignals to enable silent communication, such as electrophysiological recordings of neural activity, electromyographic (EMG) recordings of vocal tract movements or the direct tracking of articulator movements using imaging techniques. Depending on the disorder, some sensing techniques may be better suited than others to capture speech-related information. For instance, EMG and imaging techniques are well suited for laryngectomised patients, whose vocal tract remains almost intact but are unable to speak after the removal of the vocal folds, but fail for severely paralysed individuals. From the biosignals, SSIs decode the intended message, using automatic speech recognition or speech synthesis algorithms. Despite considerable advances in recent years, most present-day SSIs have only been validated in laboratory settings for healthy users. Thus, as discussed in this paper, a number of challenges remain to be addressed in future research before SSIs can be promoted to real-world applications. If these issues can be addressed successfully, future SSIs will improve the lives of persons with severe speech impairments by restoring their communication capabilities.