Showing 1–50 of 61 results for author: Miguel, A

Predefined Prototypes for Intra-Class Separation and Disentanglement

Authors: Antonio Almudévar, Théo Mariotte, Alfonso Ortega, Marie Tahon, Luis Vicente, Antonio Miguel, Eduardo Lleida

Abstract: Prototypical Learning is based on the idea that there is a point (which we call prototype) around which the embeddings of a class are clustered. It has shown promising results in scenarios with little labeled data or to design explainable models. Typically, prototypes are either defined as the average of the embeddings of a class or are designed to be trainable. In this work, we propose to predefi… ▽ More Prototypical Learning is based on the idea that there is a point (which we call prototype) around which the embeddings of a class are clustered. It has shown promising results in scenarios with little labeled data or to design explainable models. Typically, prototypes are either defined as the average of the embeddings of a class or are designed to be trainable. In this work, we propose to predefine prototypes following human-specified criteria, which simplify the training pipeline and brings different advantages. Specifically, in this work we explore two of these advantages: increasing the inter-class separability of embeddings and disentangling embeddings with respect to different variance factors, which can translate into the possibility of having explainable predictions. Finally, we propose different experiments that help to understand our proposal and demonstrate empirically the mentioned advantages. △ Less

Submitted 23 June, 2024; originally announced June 2024.

Sim-to-Real gap in RL: Use Case with TIAGo and Isaac Sim/Gym

Authors: Jaume Albardaner, Alberto San Miguel, Néstor García, Magí Dalmau-Moreno

Abstract: This paper explores policy-learning approaches in the context of sim-to-real transfer for robotic manipulation using a TIAGo mobile manipulator, focusing on two state-of-art simulators, Isaac Gym and Isaac Sim, both developed by Nvidia. Control architectures are discussed, with a particular emphasis on achieving collision-less movement in both simulation and the real environment. Presented results… ▽ More This paper explores policy-learning approaches in the context of sim-to-real transfer for robotic manipulation using a TIAGo mobile manipulator, focusing on two state-of-art simulators, Isaac Gym and Isaac Sim, both developed by Nvidia. Control architectures are discussed, with a particular emphasis on achieving collision-less movement in both simulation and the real environment. Presented results demonstrate successful sim-to-real transfer, showcasing similar movements executed by an RL-trained model in both simulated and real setups. △ Less

Submitted 27 March, 2024; v1 submitted 11 March, 2024; originally announced March 2024.

Comments: Accepted in ERF24 workshop "Towards Efficient and Portable Robot Learning for Real-World Settings". To be published in Springer Proceedings in Advanced Robotics

Assessing light pollution in vast areas: zenith sky brightness maps of Catalonia

Authors: Hector Linares, Eduard Masana Salvador J Ribas, Manuel García-Gil, Martin Aubé, Alejandro Sánchez de Miguel, Alexandre Simoneau

Abstract: Zenith sky brightness maps in the V and B bands of the region of Catalonia are presented in this paper. For creating them we have used the light pollution numerical model Illumina v2. The maps have a sampling of 5x5 km for the whole region with an improved resolution of 1x1 km for one of the provinces within Catalonia, Tarragona. Before creating the final maps, the methodology was tested successfu… ▽ More Zenith sky brightness maps in the V and B bands of the region of Catalonia are presented in this paper. For creating them we have used the light pollution numerical model Illumina v2. The maps have a sampling of 5x5 km for the whole region with an improved resolution of 1x1 km for one of the provinces within Catalonia, Tarragona. Before creating the final maps, the methodology was tested successfully by comparing the computed values to measurements in nineteen different locations spread out throughout the territory. The resulting maps have been compared to the zenith sky brightness world atlas and also to Sky Quality Meter (SQM) dynamic measurements. When comparing to measurements we found small differences mainly due to mismatching in the location of the points studied, and also due to differences in the natural sky brightness and atmospheric content. In the comparison to the world atlas some differences were expected as we are taking into account the blocking effect of topography and obstacles, and also due to a more precise light sources characterization. The results of this work confirm the conclusion found in other studies that the minimum sampling for studying sky brightness fine details is of 1x1 km. However, a sampling of 5x5 km is interesting when studying general trends, mainly for vast areas, due to the reduction of the time required to create the maps. △ Less

Submitted 29 February, 2024; originally announced March 2024.

Comments: 32 pag, 17 figures

Journal ref: Journal of Quantitative Spectroscopy and Radiative Transfer, Volume 309, 2023, 108678

Permutation Invariant Recurrent Neural Networks for Sound Source Tracking Applications

Authors: David Diaz-Guerra, Archontis Politis, Antonio Miguel, Jose R. Beltran, Tuomas Virtanen

Abstract: Many multi-source localization and tracking models based on neural networks use one or several recurrent layers at their final stages to track the movement of the sources. Conventional recurrent neural networks (RNNs), such as the long short-term memories (LSTMs) or the gated recurrent units (GRUs), take a vector as their input and use another vector to store their state. However, this approach re… ▽ More Many multi-source localization and tracking models based on neural networks use one or several recurrent layers at their final stages to track the movement of the sources. Conventional recurrent neural networks (RNNs), such as the long short-term memories (LSTMs) or the gated recurrent units (GRUs), take a vector as their input and use another vector to store their state. However, this approach results in the information from all the sources being contained in a single ordered vector, which is not optimal for permutation-invariant problems such as multi-source tracking. In this paper, we present a new recurrent architecture that uses unordered sets to represent both its input and its state and that is invariant to the permutations of the input set and equivariant to the permutations of the state set. Hence, the information of every sound source is represented in an individual embedding and the new estimates are assigned to the tracked trajectories regardless of their order. △ Less

Submitted 14 June, 2023; originally announced June 2023.

Comments: Accepted for publication at Forum Acusticum 2023

Characterisation of night-time outdoor lighting in urban centres using cluster analysis of remotely sensed light emissions

Authors: Máximo Bustamante-Calabria, Susana Martín-Ruiz, Alejandro Sánchez de Miguel, J. L. Ortiz, J. M. Vílchez, Jesús Aceituno

Abstract: Evidence of the negative impact of light pollution on ecosystems is increasing every year. Its monitoring and study requires the identification, characterisation and control of the emitting sources. This is the case of urban centres with outdoor lighting that spills light outside the place it is intended to illuminate. The quantity and nature of the pollutant (artificial light at night) depends on… ▽ More Evidence of the negative impact of light pollution on ecosystems is increasing every year. Its monitoring and study requires the identification, characterisation and control of the emitting sources. This is the case of urban centres with outdoor lighting that spills light outside the place it is intended to illuminate. The quantity and nature of the pollutant (artificial light at night) depends on the lamps used and how they are positioned. This is important because a greater proportion of blue light means a greater scattering effect. In this study, we analysed the emissions of 100 urban centres in the north of Granada province (Spain), using International Space Station (ISS) images from 2012 and 2021, in order to compare the results with public lighting inventories and verify the validity of these data for characterising night-time lighting emissions. Using inference and cluster analysis techniques, we confirmed an overall increase in emissions and a shift in their colour towards blue, consistent with the results of the lighting inventory analysis. We concluded that it is possible to use ISS imagery to characterise artificial light emissions and the lighting that causes them, none the less there are a number of inherent problems with the data and the way it was collected that require the results to be interpreted with caution. △ Less

Submitted 31 December, 2023; v1 submitted 19 May, 2023; originally announced May 2023.

Photometric Catalogue for Space and Ground Night-Time Remote-Sensing Calibration: RGB Synthetic Photometry from Gaia DR3 Spectrophotometry

Authors: J. M. Carrasco, N. Cardiel, E. Masana, J. Zamorano, S. Pascual, A. Sánchez de Miguel, R. González, J. Izquierdo

Abstract: Recent works have made strong efforts to produce standardised photometry in RGB bands. For this purpose, we carefully defined the transmissivity curves of RGB bands and defined a set of standard sources using the photometric information present in Gaia EDR3. This work aims not only to significantly increase the number and accuracy of RGB standards but also to provide, for the first time, reliable… ▽ More Recent works have made strong efforts to produce standardised photometry in RGB bands. For this purpose, we carefully defined the transmissivity curves of RGB bands and defined a set of standard sources using the photometric information present in Gaia EDR3. This work aims not only to significantly increase the number and accuracy of RGB standards but also to provide, for the first time, reliable uncertainty estimates using the BP and RP spectrophotometry published in Gaia DR3 instead of their integrated photometry to predict RGB photometry. Furthermore, this method allows including calibrated sources regardless of how they are affected by extinction, which was a major shortcoming of previous work. The RGB photometry is synthesised from the Gaia BP and RP low-resolution spectra by directly using their set of coefficients multiplied with some basis functions provided in the Gaia catalogue for all sources published in Gaia DR3. The output synthetic magnitudes are compared with the previous catalogue of RGB standards available. △ Less

Submitted 24 March, 2023; originally announced March 2023.

Comments: 23 pages, 15 figures

Importance of the time acquisition difference between DMSP/OLS and SNPP/VIIRS/DNB and the RAW trends in Europe

Authors: Alejandro Sánchez de Miguel, Sara Krupansky

Abstract: Using the SNPP-VIIRS/DNB and the DMSP-OLS it is possible to have an idea of the evolution of the light pollution until the LEDs started to appear massively. Another of the issues of dealing with these two datasets is the different time of acquisition. In general this means that countries get dimmer late at night, although that is not always true. A counterexample of this is the country of the Neth… ▽ More Using the SNPP-VIIRS/DNB and the DMSP-OLS it is possible to have an idea of the evolution of the light pollution until the LEDs started to appear massively. Another of the issues of dealing with these two datasets is the different time of acquisition. In general this means that countries get dimmer late at night, although that is not always true. A counterexample of this is the country of the Netherlands because of the opening of the shutters late at night. Other countries have a much bigger drop late at night because of some practices like the total or partial turn off of public and private lighting but also dimming policies. Examples of this are countries like France, Austria and the UK. △ Less

Submitted 14 October, 2022; originally announced November 2022.

CobotTouch: AR-based Interface with Fingertip-worn Tactile Display for Immersive Operation/Control of Collaborative Robots

Authors: Sautenkov Oleg, Altamirano Cabrera Miguel, Rakhmatulin Viktor, Tsetserukou Dzmitry

Abstract: Complex robotic tasks require human collaboration to benefit from their high dexterity. Frequent human-robot interaction is mentally demanding and time-consuming. Intuitive and easy-to-use robot control interfaces reduce the negative influence on workers, especially inexperienced users. In this paper, we present CobotTouch, a novel intuitive robot control interface with fingertip haptic feedback.… ▽ More Complex robotic tasks require human collaboration to benefit from their high dexterity. Frequent human-robot interaction is mentally demanding and time-consuming. Intuitive and easy-to-use robot control interfaces reduce the negative influence on workers, especially inexperienced users. In this paper, we present CobotTouch, a novel intuitive robot control interface with fingertip haptic feedback. The proposed interface consists of a projected Graphical User Interface on the robotic arm to control the position of the robot end-effector based on gesture recognition, and a wearable haptic interface to deliver tactile feedback on the user's fingertips. We evaluated the user's perception of the designed tactile patterns presented by the haptic interface and the intuitiveness of the proposed system for robot control in a use case. The results revealed a high average recognition rate of 75.25\% for the tactile patterns. An average NASA Task Load Index (TLX) indicated small mental and temporal demands proving a high level of the intuitiveness of CobotTouch for interaction with collaborative robots. △ Less

Submitted 2 November, 2022; originally announced November 2022.

Comments: 12 pages, 11 figures, Accepted paper in AsiaHaptics 2022

Citizen Science to Assess Light Pollution with Mobile Phones

Authors: Gorka Muñoz-Gil, Alexandre Dauphin, Federica A. Beduini, Alejandro Sánchez de Miguel

Abstract: The analysis of the colour of artificial lights at night has an impact on diverse fields, but current data sources have either limited resolution or scarce availability of images for a specific region. In this work, we propose crowdsourced photos of streetlights as an alternative data source: for this, we designed NightUp Castelldefels, a pilot for a citizen science experiment aimed at collecting… ▽ More The analysis of the colour of artificial lights at night has an impact on diverse fields, but current data sources have either limited resolution or scarce availability of images for a specific region. In this work, we propose crowdsourced photos of streetlights as an alternative data source: for this, we designed NightUp Castelldefels, a pilot for a citizen science experiment aimed at collecting data about the colour of streetlights. In particular, we extract the colour from the collected images and compare it to an official database, showing that it is possible to classify streetlights according to their colour from photos taken by untrained citizens with their own smartphones. We also compare our findings to the results obtained from one of the current sources for this kind of study. The comparison highlights how the two approaches give complementary information about artificial lights at night in the area. This work opens a new avenue in the study of the colour of artificial lights at night with the possibility of accurate, massive and cheap data collection. △ Less

Submitted 20 October, 2022; originally announced October 2022.

Journal ref: Remote Sensing, 14(19), 4976

Non-universality in clustered ballistic annihilation

Authors: Matthew Junge, Arturo Ortiz San Miguel, Lily Reeves, Cynthia Rivera Sánchez

Abstract: In ballistic annihilation, infinitely many particles with randomly assigned velocities move across the real line and mutually annihilate upon contact. We introduce a variant with superimposed clusters of multiple stationary particles. Our main result is that the critical initial cluster density to ensure species survival depends on both the mean and variance of the cluster size. Our result contras… ▽ More In ballistic annihilation, infinitely many particles with randomly assigned velocities move across the real line and mutually annihilate upon contact. We introduce a variant with superimposed clusters of multiple stationary particles. Our main result is that the critical initial cluster density to ensure species survival depends on both the mean and variance of the cluster size. Our result contrasts with recent ballistic annihilation universality findings with respect to particle spacings. A corollary of our theorem resolves an open question for coalescing ballistic annihilation. △ Less

Submitted 23 September, 2022; v1 submitted 9 September, 2022; originally announced September 2022.

Comments: 15 pages, 1 figure

MSC Class: 60K35

Pros and cons of gaussian filters versus step filters for light pollution monitoring

Authors: Alejandro Sánchez de Miguel

Abstract: There is debate about which indicators should currently be used to monitor levels of artificial light pollution. To be most valuable, methods need to be sensitive to variation in the spectral composition of light emissions (which are changing rapidly, particularly through increasing use of light-emitting diode [LED] lamps), to be readily available, to be capable of being used on a large spatial sc… ▽ More There is debate about which indicators should currently be used to monitor levels of artificial light pollution. To be most valuable, methods need to be sensitive to variation in the spectral composition of light emissions (which are changing rapidly, particularly through increasing use of light-emitting diode [LED] lamps), to be readily available, to be capable of being used on a large spatial scale and of being deployed rapidly. Two sets of photometric systems are the most spread in the world currently, the RGB colors from DSLR cameras that are based on typical gaussian filters and RGB step filters. The first set of filters are optimum for human perception and calculation of most of the most popular environmental impacts although, some of these environmental impacts are better characterized by the step filters. △ Less

Submitted 5 September, 2022; originally announced September 2022.

Arigatō: Effects of Adaptive Guidance on Engagement and Performance in Augmented Reality Learning Environments

Authors: Maheshya Weerasinghe, Aaron Quigley, Klen Čopič Pucihar, Alice Toniolo, Angela Miguel, Matjaž Kljun

Abstract: Experiential learning (ExL) is the process of learning through experience or more specifically "learning through reflection on doing". In this paper, we propose a simulation of these experiences, in Augmented Reality (AR), addressing the problem of language learning. Such systems provide an excellent setting to support "adaptive guidance", in a digital form, within a real environment. Adaptive gui… ▽ More Experiential learning (ExL) is the process of learning through experience or more specifically "learning through reflection on doing". In this paper, we propose a simulation of these experiences, in Augmented Reality (AR), addressing the problem of language learning. Such systems provide an excellent setting to support "adaptive guidance", in a digital form, within a real environment. Adaptive guidance allows the instructions and learning content to be customised for the individual learner, thus creating a unique learning experience. We developed an adaptive guidance AR system for language learning, we call Arigatō (Augmented Reality Instructional Guidance & Tailored Omniverse), which offers immediate assistance, resources specific to the learner's needs, manipulation of these resources, and relevant feedback. Considering guidance, we employ this prototype to investigate the effect of the amount of guidance (fixed vs. adaptive-amount) and the type of guidance (fixed vs. adaptive-associations) on the engagement and consequently the learning outcomes of language learning in an AR environment. The results for the amount of guidance show that compared to the adaptive-amount, the fixed-amount of guidance group scored better in the immediate and delayed (after 7 days) recall tests. However, this group also invested a significantly higher mental effort to complete the task. The results for the type of guidance show that the adaptive-associations group outperforms the fixed-associations group in the immediate, delayed (after 7 days) recall tests, and learning efficiency. The adaptive-associations group also showed significantly lower mental effort and spent less time to complete the task. △ Less

Submitted 2 July, 2022; originally announced July 2022.

Comments: 11 pages, 2 figures, 2 tables

ACM Class: I.3.7

DeepXPalm: Tilt and Position Rendering using Palm-worn Haptic Display and CNN-based Tactile Pattern Recognition

Authors: Altamirano Cabrera Miguel, Sautenkov Oleg, Tirado Jonathan, Fedoseev Aleksey, Kopanev Pavel, Kajimoto Hiroyuki, Tsetserukou Dzmitry

Abstract: Telemanipulation of deformable objects requires high precision and dexterity from the users, which can be increased by kinesthetic and tactile feedback. However, the object shape can change dynamically, causing ambiguous perception of its alignment and hence errors in the robot positioning. Therefore, the tilt angle and position classification problem has to be solved to present a clear tactile pa… ▽ More Telemanipulation of deformable objects requires high precision and dexterity from the users, which can be increased by kinesthetic and tactile feedback. However, the object shape can change dynamically, causing ambiguous perception of its alignment and hence errors in the robot positioning. Therefore, the tilt angle and position classification problem has to be solved to present a clear tactile pattern to the user. This work presents a telemanipulation system for plastic pipettes consisting of a multi-contact haptic device LinkGlide to deliver haptic feedback at the users' palm and two tactile sensors array embedded in the 2-finger Robotiq gripper. We propose a novel approach based on Convolutional Neural Networks (CNN) to detect the tilt and position while gras** deformable objects. The CNN generates a mask based on recognized tilt and position data to render further multi-contact tactile stimuli provided to the user during the telemanipulation. The study has shown that using the CNN algorithm and the preset mask, tilt, and position recognition by users is increased from 9.67% using the direct data to 82.5%. △ Less

Submitted 7 April, 2022; originally announced April 2022.

Comments: Accepted paper in IEEE Haptic Symposium 2022, IEEE copyright

Direction of Arrival Estimation of Sound Sources Using Icosahedral CNNs

Authors: David Diaz-Guerra, Antonio Miguel, Jose R. Beltran

Abstract: In this paper, we present a new model for Direction of Arrival (DOA) estimation of sound sources based on an Icosahedral Convolutional Neural Network (CNN) applied over SRP-PHAT power maps computed from the signals received by a microphone array. This icosahedral CNN is equivariant to the 60 rotational symmetries of the icosahedron, which represent a good approximation of the continuous space of s… ▽ More In this paper, we present a new model for Direction of Arrival (DOA) estimation of sound sources based on an Icosahedral Convolutional Neural Network (CNN) applied over SRP-PHAT power maps computed from the signals received by a microphone array. This icosahedral CNN is equivariant to the 60 rotational symmetries of the icosahedron, which represent a good approximation of the continuous space of spherical rotations, and can be implemented using standard 2D convolutional layers, having a lower computational cost than most of the spherical CNNs. In addition, instead of using fully connected layers after the icosahedral convolutions, we propose a new soft-argmax function that can be seen as a differentiable version of the argmax function and allows us to solve the DOA estimation as a regression problem interpreting the output of the convolutional layers as a probability distribution. We prove that using models that fit the equivariances of the problem allows us to outperform other state-of-the-art models with a lower computational cost and more robustness, obtaining root mean square localization errors lower than 10° even in scenarios with a reverberation time $T_{60}$ of 1.5 s. △ Less

Submitted 6 December, 2022; v1 submitted 31 March, 2022; originally announced March 2022.

Comments: The code to reproduce this work can be found in our GitHub repository: https://github.com/DavidDiazGuerra/icoDOA

Journal ref: IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 31, pp. 313-321, 2023

Class Token and Knowledge Distillation for Multi-head Self-Attention Speaker Verification Systems

Authors: Victoria Mingote, Antonio Miguel, Alfonso Ortega, Eduardo Lleida

Abstract: This paper explores three novel approaches to improve the performance of speaker verification (SV) systems based on deep neural networks (DNN) using Multi-head Self-Attention (MSA) mechanisms and memory layers. Firstly, we propose the use of a learnable vector called Class token to replace the average global pooling mechanism to extract the embeddings. Unlike global average pooling, our proposal t… ▽ More This paper explores three novel approaches to improve the performance of speaker verification (SV) systems based on deep neural networks (DNN) using Multi-head Self-Attention (MSA) mechanisms and memory layers. Firstly, we propose the use of a learnable vector called Class token to replace the average global pooling mechanism to extract the embeddings. Unlike global average pooling, our proposal takes into account the temporal structure of the input what is relevant for the text-dependent SV task. The class token is concatenated to the input before the first MSA layer, and its state at the output is used to predict the classes. To gain additional robustness, we introduce two approaches. First, we have developed a Bayesian estimation of the class token. Second, we have added a distilled representation token for training a teacher-student pair of networks using the Knowledge Distillation (KD) philosophy, which is combined with the class token. This distillation token is trained to mimic the predictions from the teacher network, while the class token replicates the true label. All the strategies have been tested on the RSR2015-Part II and DeepMine-Part 1 databases for text-dependent SV, providing competitive results compared to the same architecture using the average pooling mechanism to extract average embeddings. △ Less

Submitted 10 February, 2023; v1 submitted 6 November, 2021; originally announced November 2021.

Generalizing AUC Optimization to Multiclass Classification for Audio Segmentation With Limited Training Data

Authors: Pablo Gimeno, Victoria Mingote, Alfonso Ortega, Antonio Miguel, Eduardo Lleida

Abstract: Area under the ROC curve (AUC) optimisation techniques developed for neural networks have recently demonstrated their capabilities in different audio and speech related tasks. However, due to its intrinsic nature, AUC optimisation has focused only on binary tasks so far. In this paper, we introduce an extension to the AUC optimisation framework so that it can be easily applied to an arbitrary numb… ▽ More Area under the ROC curve (AUC) optimisation techniques developed for neural networks have recently demonstrated their capabilities in different audio and speech related tasks. However, due to its intrinsic nature, AUC optimisation has focused only on binary tasks so far. In this paper, we introduce an extension to the AUC optimisation framework so that it can be easily applied to an arbitrary number of classes, aiming to overcome the issues derived from training data limitations in deep learning solutions. Building upon the multiclass definitions of the AUC metric found in the literature, we define two new training objectives using a one-versus-one and a one-versus-rest approach. In order to demonstrate its potential, we apply them in an audio segmentation task with limited training data that aims to differentiate 3 classes: foreground music, background music and no music. Experimental results show that our proposal can improve the performance of audio segmentation systems significantly compared to traditional training criteria such as cross entropy. △ Less

Submitted 27 October, 2021; originally announced October 2021.

Journal ref: IEEE Signal Processing Letters, vol. 28, pp. 1135-1139, 2021

CobotAR: Interaction with Robots using Omnidirectionally Projected Image and DNN-based Gesture Recognition

Authors: Nazarova Elena, Sautenkov Oleg, Altamirano Cabrera Miguel, Tirado Jonathan, Serpiva Valerii, Rakhmatulin Viktor, Tsetserukou Dzmitry

Abstract: Several technological solutions supported the creation of interfaces for Augmented Reality (AR) multi-user collaboration in the last years. However, these technologies require the use of wearable devices. We present CobotAR - a new AR technology to achieve the Human-Robot Interaction (HRI) by gesture recognition based on Deep Neural Network (DNN) - without an extra wearable device for the user. Th… ▽ More Several technological solutions supported the creation of interfaces for Augmented Reality (AR) multi-user collaboration in the last years. However, these technologies require the use of wearable devices. We present CobotAR - a new AR technology to achieve the Human-Robot Interaction (HRI) by gesture recognition based on Deep Neural Network (DNN) - without an extra wearable device for the user. The system allows users to have a more intuitive experience with robotic applications using just their hands. The CobotAR system assumes the AR spatial display created by a mobile projector mounted on a 6 DoF robot. The proposed technology suggests a novel way of interaction with machines to achieve safe, intuitive, and immersive control mediated by a robotic projection system and DNN-based algorithm. We conducted the experiment with several parameters assessment during this research, which allows the users to define the positives and negatives of the new approach. The mental demand of CobotAR system is twice less than Wireless Gamepad and by 16\% less than Teach Pendant. △ Less

Submitted 20 October, 2021; originally announced October 2021.

Comments: Accepted paper in SMC conference 2021, IEEE copyright

Colour remote sensing of the impact of artificial light at night (II): Calibration of DSLR-based images from the International Space Station

Authors: Alejandro Sánchez de Miguel, Jaime Zamorano, Martin Aubé, Jonathan Bennie, Jesús Gallego, Francisco Ocaña, Donald R. Pettit, William L. Stefanov, Kevin J. Gaston

Abstract: Nighttime images taken with DSLR cameras from the International Space Station (ISS) can provide valuable information on the spatial and temporal variation of artificial nighttime lighting on Earth. In particular, this is the only source of historical and current visible multispectral data across the world (DMSP/OLS and SNPP/VIIRS-DNB data are panchromatic and multispectral in the infrared but not… ▽ More Nighttime images taken with DSLR cameras from the International Space Station (ISS) can provide valuable information on the spatial and temporal variation of artificial nighttime lighting on Earth. In particular, this is the only source of historical and current visible multispectral data across the world (DMSP/OLS and SNPP/VIIRS-DNB data are panchromatic and multispectral in the infrared but not at visible wavelengths). The ISS images require substantial processing and proper calibration to exploit intensities and ratios from the RGB channels. Here we describe the different calibration steps, addressing in turn Decodification, Linearity correction (ISO dependent), Flat field/Vignetting, Spectral characterization of the channels, Astrometric calibration/georeferencing, Photometric calibration (stars)/Radiometric correction (settings correction - by exposure time, ISO, lens transmittance, etc) and Transmittance correction (window transmittance, atmospheric correction). We provide an example of the application of this processing method to an image of Spain. △ Less

Submitted 16 August, 2021; originally announced August 2021.

Journal ref: Sánchez de Miguel et. al. (2021). Colour remote sensing of the impact of artificial light at night (II): Calibration of DSLR-based images from the International Space Station. Remote Sensing of Environment, 112611

RGB photometric calibration of 15 million Gaia stars

Authors: Nicolás Cardiel, Jaime Zamorano, Josep Manel Carrasco, Eduard Masana, Salvador Bará, Rafael González, Jaime Izquierdo, Sergio Pascual, Alejandro Sánchez de Miguel

Abstract: Although a catalogue of synthetic RGB magnitudes, providing photometric data for a sample of 1346 bright stars, has been recently published, its usefulness is still limited due to the small number of reference stars available, considering that they are distributed throughout the whole celestial sphere, and the fact that they are restricted to Johnson V < 6.6 mag. This work presents synthetic RGB m… ▽ More Although a catalogue of synthetic RGB magnitudes, providing photometric data for a sample of 1346 bright stars, has been recently published, its usefulness is still limited due to the small number of reference stars available, considering that they are distributed throughout the whole celestial sphere, and the fact that they are restricted to Johnson V < 6.6 mag. This work presents synthetic RGB magnitudes for ~15 million stars brighter than Gaia G = 18 mag, making use of a calibration between the RGB magnitudes of the reference bright star sample and the corresponding high quality photometric G, G_BP and G_RP magnitudes provided by the Gaia EDR3. The calibration has been restricted to stars exhibiting -0.5 < G_BP - G_RP < 2.0 mag, and aims to predict RGB magnitudes within an error interval of $\pm 0.1$ mag. Since the reference bright star sample is dominated by nearby stars with slightly undersolar metallicity, systematic variations in the predictions are expected, as modelled with the help of stellar atmosphere models. These deviations are constrained to the $\pm 0.1$ mag interval when applying the calibration only to stars scarcely affected by interstellar extinction and with metallicity compatible with the median value for the bright star sample. The large number of Gaia sources available in each region of the sky should guarantee high-quality RGB photometric calibrations. △ Less

Submitted 19 July, 2021; originally announced July 2021.

Comments: 12 pages, 9 figures, 2 tables. Accepted for publication in MNRAS

Synthetic RGB photometry of bright stars: definition of the standard photometric system and UCM library of spectrophotometric spectra

Authors: Nicolás Cardiel, Jaime Zamorano, Salvador Bará, Alejandro Sánchez de Miguel, Cristina Cabello, Jesús Gallego, Lucía García, Rafael González, Jaime Izquierdo, Sergio Pascual, José Robles, Ainhoa Sánchez, Carlos Tapia

Abstract: Although the use of RGB photometry has exploded in the last decades due to the advent of high-quality and inexpensive digital cameras equipped with Bayer-like color filter systems, there is surprisingly no catalogue of bright stars that can be used for calibration purposes. Since due to their excessive brightness, accurate enough spectrophotometric measurements of bright stars typically cannot be… ▽ More Although the use of RGB photometry has exploded in the last decades due to the advent of high-quality and inexpensive digital cameras equipped with Bayer-like color filter systems, there is surprisingly no catalogue of bright stars that can be used for calibration purposes. Since due to their excessive brightness, accurate enough spectrophotometric measurements of bright stars typically cannot be performed with modern large telescopes, we have employed historical 13-color medium-narrow-band photometric data, gathered with quite reliable photomultipliers, to fit the spectrum of 1346 bright stars using stellar atmosphere models. This not only constitutes a useful compilation of bright spectrophotometric standards well spread in the celestial sphere, the UCM library of spectrophotometric spectra, but allows the generation of a catalogue of reference RGB magnitudes, with typical random uncertainties $\sim 0.01$ mag. For that purpose, we have defined a new set of spectral sensitivity curves, computed as the median of 28 sets of empirical sensitivity curves from the literature, that can be used to establish a standard RGB photometric system. Conversions between RGB magnitudes computed with any of these sets of empirical RGB curves and those determined with the new standard photometric system are provided. Even though particular RGB measurements from single cameras are not expected to provide extremely accurate photometric data, the repeatability and multiplicity of observations will allow access to a large amount of exploitable data in many astronomical fields, such as the detailed monitoring of light pollution and its impact on the night sky brightness, or the study of meteors, solar system bodies, variable stars, and transient objects. In addition, the RGB magnitudes presented here make the sky an accessible and free laboratory for the calibration of the cameras themselves. △ Less

Submitted 31 March, 2021; originally announced March 2021.

Comments: 19 pages, 18 figures, 7 tables. Accepted for publication in MNRAS

Direct assessment of the sensitivity drift of SQM sensors installed outdoors

Authors: Salvador Bará, Enric Marco, Salvador J. Ribas, Manuel Garcia Gil, Alejandro Sánchez de Miguel, Jaime Zamorano

Abstract: Long-term monitoring of the evolution of the artificial night sky brightness is a key tool for develo** science-informed public policies and assessing the efficacy of light pollution mitigation measures. Detecting the underlying artificial brightness trend is a challenging task, since the typical night sky brightness signal shows a large variability with characteristic time scales ranging from s… ▽ More Long-term monitoring of the evolution of the artificial night sky brightness is a key tool for develo** science-informed public policies and assessing the efficacy of light pollution mitigation measures. Detecting the underlying artificial brightness trend is a challenging task, since the typical night sky brightness signal shows a large variability with characteristic time scales ranging from seconds to years. In order to effectively isolate the weak signature of the effect of interest, determining the potential long term drifts of the radiance sensing systems is crucial. If these drifts can be adequately characterized, the raw measurements could be easily corrected for them and transformed to a consistent scale. In this short note we report on the progressive darkening of the signal recorded by SQM detectors belonging to several monitoring networks, permanently installed outdoors for periods ranging from several months to several years. The sensitivity drifts were estimated by means of parallel measurements made at the beginning and at the end of the evaluation periods using reference detectors of the same kind that were little or no exposed to weathering in the intervening time. Our preliminary results suggest that SQM detectors installed outdoors steadily increase their readings at an average rate of +0.034 magSQM/arcsec^2 per MWh/m^2 of exposure to solar horizontal global irradiation, that for our locations translates into approximately +0.05 to +0.06 magSQM/arcsec^2 per year. △ Less

Submitted 1 December, 2020; v1 submitted 30 November, 2020; originally announced November 2020.

Comments: 5 pages, 1 figure. ArXiv identifier added, some minor misprints corrected

Softening of the Euler buckling criterion under discretisation of compliance

Authors: D. J. Carter, D. J. Dunstan, W. Just, O. F. Bandtlow, A. San Miguel

Abstract: Euler solved the problem of the collapse of tall thin columns under unexpectedly small loads in 1744. The analogous problem of the collapse of circular elastic rings or tubes under external pressure was mathematically intractable and only fully solved recently. In the context of carbon nanotubes, an additional phenomenon was found experimentally and in atomistic simulations but not explained: the… ▽ More Euler solved the problem of the collapse of tall thin columns under unexpectedly small loads in 1744. The analogous problem of the collapse of circular elastic rings or tubes under external pressure was mathematically intractable and only fully solved recently. In the context of carbon nanotubes, an additional phenomenon was found experimentally and in atomistic simulations but not explained: the collapse pressure of smaller diameter tubes deviates below the continuum mechanics solution [Torres-Dias et al., Carbon 123, 145 (2017)]. Here, this deviation is shown to occur in discretized straight columns and it is fully explained in terms of the phonon dispersion curve. This reveals an unexpected link between the static mechanical properties of discrete systems and their dynamics described through dispersion curves. △ Less

Submitted 28 November, 2020; originally announced November 2020.

Comments: 4 pages, 3 Figures, draft in preparation

Effects of the COVID-19 lockdown on urban light emissions: ground and satellite comparison

Authors: Máximo Bustamante-Calabria, Alejandro Sánchez de Miguel, Susana Martín-Ruiz, Jose-Luis Ortiz, J. M. Vílchez, Alicia Pelegrina, Antonio García, Jaime Zamorano, Jonathan Bennie, Kevin J. Gaston

Abstract: 'Lockdown' periods in response to COVID-19 have provided a unique opportunity to study the impacts of economic activity on environmental pollution (e.g. NO$_2$, aerosols, noise, light). The effects on NO$_2$ and aerosols have been very noticeable and readily demonstrated, but that on light pollution has proven challenging to determine. The main reason for this difficulty is that the primary source… ▽ More 'Lockdown' periods in response to COVID-19 have provided a unique opportunity to study the impacts of economic activity on environmental pollution (e.g. NO$_2$, aerosols, noise, light). The effects on NO$_2$ and aerosols have been very noticeable and readily demonstrated, but that on light pollution has proven challenging to determine. The main reason for this difficulty is that the primary source of nighttime satellite imagery of the earth is the SNPP-VIIRS/DNB instrument, which acquires data late at night after most human nocturnal activity has already occurred and much associated lighting has been turned off. Here, to analyze the effect of lockdown on urban light emissions, we use ground and satellite data for Granada, Spain, during the COVID-19 induced confinement of the city's population from March 14 until May 31, 2020. We find a clear decrease in light pollution due both to a decrease in light emissions from the city and to a decrease in anthropogenic aerosol content in the atmosphere which resulted in less light being scattered. A clear correlation between the abundance of PM10 particles and sky brightness is observed, such that the more polluted the atmosphere the brighter the urban night sky. An empirical expression is determined that relates PM10 particle abundance and sky brightness at three different wavelength bands. △ Less

Submitted 18 November, 2020; originally announced November 2020.

Journal ref: Remote Sens. 2021, 13(2), 258

Supervised learning of few dirty bosons with variable particle number

Authors: Pere Mujal, Àlex Martínez Miguel, Artur Polls, Bruno Juliá-Díaz, Sebastiano Pilati

Abstract: We investigate the supervised machine learning of few interacting bosons in optical speckle disorder via artificial neural networks. The learning curve shows an approximately universal power-law scaling for different particle numbers and for different interaction strengths. We introduce a network architecture that can be trained and tested on heterogeneous datasets including different particle num… ▽ More We investigate the supervised machine learning of few interacting bosons in optical speckle disorder via artificial neural networks. The learning curve shows an approximately universal power-law scaling for different particle numbers and for different interaction strengths. We introduce a network architecture that can be trained and tested on heterogeneous datasets including different particle numbers. This network provides accurate predictions for all system sizes included in the training set and, by design, is suitable to attempt extrapolations to (computationally challenging) larger sizes. Notably, a novel transfer-learning strategy is implemented, whereby the learning of the larger systems is substantially accelerated and made consistently accurate by including in the training set many small-size instances. △ Less

Submitted 28 February, 2021; v1 submitted 8 October, 2020; originally announced October 2020.

Comments: Submission to SciPost

Journal ref: SciPost Phys. 10, 073 (2021)

Application of Statistical Methods in Software Engineering: Theory and Practice

Authors: T. F. M. Sirqueira, M. A. Miguel, H. L. O. Dalpra, M. A. P. Araujo, J. M. N. David

Abstract: The experimental evaluation of the methods and concepts covered in software engineering has been increasingly valued. This value indicates the constant search for new forms of assessment and validation of the results obtained in Software Engineering research. Results are validated in studies through evaluations, which in turn become increasingly stringent. As an alternative to aid in the verificat… ▽ More The experimental evaluation of the methods and concepts covered in software engineering has been increasingly valued. This value indicates the constant search for new forms of assessment and validation of the results obtained in Software Engineering research. Results are validated in studies through evaluations, which in turn become increasingly stringent. As an alternative to aid in the verification of the results, that is, whether they are positive or negative, we suggest the use of statistical methods. This article presents some of the main statistical techniques available, as well as their use in carrying out the implementation of data analysis in experimental studies in Software Engineering. This paper presents a practical approach proving statistical techniques through a decision tree, which was created in order to facilitate the understanding of the appropriate statistical method for each data analysis situation. Actual data from the software projects were employed to demonstrate the use of these statistical methods. Although it is not the aim of this work, basic experimentation and statistics concepts will be presented, as well as a concrete indication of the applicability of these techniques. △ Less

Submitted 28 June, 2020; originally announced June 2020.

Response of Vulnerable Road Users to Visual Information from Autonomous Vehicles in Shared Spaces

Authors: Walter Morales Alvarez, Miguel Ángel de Miguel, Fernando García, Cristina Olaverri-Monreal

Abstract: Completely unmanned autonomous vehicles have been anticipated for a while. Initially, these are expected to drive only under certain conditions on some roads, and advanced functionality is required to cope with the ever-increasing challenges of safety. To enhance the public's perception of road safety and trust in new vehicular technologies, we investigate in this paper the effect of several inter… ▽ More Completely unmanned autonomous vehicles have been anticipated for a while. Initially, these are expected to drive only under certain conditions on some roads, and advanced functionality is required to cope with the ever-increasing challenges of safety. To enhance the public's perception of road safety and trust in new vehicular technologies, we investigate in this paper the effect of several interaction paradigms with vulnerable road users by develo** and applying algorithms for the automatic analysis of pedestrian body language. We assess behavioral patterns and determine the impact of the coexistence of AVs and other road users on general road safety in a shared space for VRUs and vehicles. Results showed that the implementation of visual communication cues for interacting with VRUs is not necessarily required for a shared space in which informal traffic rules apply. △ Less

Submitted 22 July, 2020; v1 submitted 16 June, 2020; originally announced June 2020.

Comments: Published paper in the IEEE Intelligent Transportation Systems Conference - ITSC 2019

Journal ref: 2019 IEEE Intelligent Transportation Systems Conference (ITSC), Auckland, New Zealand, 2019, pp. 3714-3719

Robust Sound Source Tracking Using SRP-PHAT and 3D Convolutional Neural Networks

Authors: David Diaz-Guerra, Antonio Miguel, Jose R. Beltran

Abstract: In this paper, we present a new single sound source DOA estimation and tracking system based on the well-known SRP-PHAT algorithm and a three-dimensional Convolutional Neural Network. It uses SRP-PHAT power maps as input features of a fully convolutional causal architecture that uses 3D convolutional layers to accurately perform the tracking of a sound source even in highly reverberant scenarios w… ▽ More In this paper, we present a new single sound source DOA estimation and tracking system based on the well-known SRP-PHAT algorithm and a three-dimensional Convolutional Neural Network. It uses SRP-PHAT power maps as input features of a fully convolutional causal architecture that uses 3D convolutional layers to accurately perform the tracking of a sound source even in highly reverberant scenarios where most of the state of the art techniques fail. Unlike previous methods, since we do not use bidirectional recurrent layers and all our convolutional layers are causal in the time dimension, our system is feasible for real-time applications and it provides a new DOA estimation for each new SRP-PHAT map. To train the model, we introduce a new procedure to simulate random trajectories as they are needed during the training, equivalent to an infinite-size dataset with high flexibility to modify its acoustical conditions such as the reverberation time. We use both acoustical simulations on a large range of reverberation times and the actual recordings of the LOCATA dataset to prove the robustness of our system and its good performance even using low-resolution SRP-PHAT maps. △ Less

Submitted 16 December, 2020; v1 submitted 16 June, 2020; originally announced June 2020.

Comments: This is a pre-print of an article published in IEEE/ACM Transactions on Audio Speech and Language Processing. The code to reproduce this work can be found in our GitHub repository: https://github.com/DavidDiazGuerra/Cross3D

Journal ref: in IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 29, pp. 300-311, 2021

Balloon-borne video observations of Geminids 2016

Authors: Francisco Ocaña, Alejandro Sánchez de Miguel, ORISON team, Daedalus Project

Abstract: We investigate the observation of meteors with video cameras in stratospheric balloons, overcoming tropospheric handicaps like weather and extinction. We have studied the practical implementation of the idea, designed and tested instrumentation for balloon-borne missions. We have analysed the data of the Geminids 2016 campaign, determining the meteoroid flux just before the maximum. This text is… ▽ More We investigate the observation of meteors with video cameras in stratospheric balloons, overcoming tropospheric handicaps like weather and extinction. We have studied the practical implementation of the idea, designed and tested instrumentation for balloon-borne missions. We have analysed the data of the Geminids 2016 campaign, determining the meteoroid flux just before the maximum. This text is an adaption of the work by the first author for his PhD Thesis: Techniques for near-Earth interplanetary matter detection and characterisation from optical ground-based observatories (Ocaña, 2017). Refer to his thesis for further detail. The lines here are a summary of the presentation given, for the sake of completeness of these proceedings of the IMC 2018 in Pezinok-Modra. The multimedia material shown during the presentation at IMC 2018 can be found in the Zenodo repository for the ORISON Project and Daedalus Project. △ Less

Submitted 12 November, 2019; originally announced November 2019.

Comments: International Meteor Conference 2018, held in Pezinok-Modra, Slovakia

Journal ref: Proceedings of the International Meteor Conference, Pezinok-Modra, Slovakia 2018 ISBN 978-2-87355-032-5

The nature of the diffuse light near cities detected in nighttime satellite imagery

Authors: Alejandro Sánchez de Miguel, Christopher C. M. Kyba, Jaime Zamorano, Jesús Gallego, Kevin J. Gaston

Abstract: Diffuse glow has been observed around brightly lit cities in nighttime satellite imagery since at least the first publication of large scale maps in the late 1990s. In the literature, this has often been assumed to be an error related to the sensor, and referred to as "blooming", presumably in relation to the effect that can occur when using a CCD to photograph a bright source. Here we show that t… ▽ More Diffuse glow has been observed around brightly lit cities in nighttime satellite imagery since at least the first publication of large scale maps in the late 1990s. In the literature, this has often been assumed to be an error related to the sensor, and referred to as "blooming", presumably in relation to the effect that can occur when using a CCD to photograph a bright source. Here we show that the effect is not instrumental, but in fact represents a real detection of light scattered by the atmosphere. Data from the Universidad Complutense Madrid sky brightness survey are compared to nighttime imagery from multiple sensors with differing spatial resolutions, and found to be strongly correlated. These results suggest that it should be possible for a future space-based imaging radiometer to monitor changes in the diffuse artificial skyglow of cities. △ Less

Submitted 15 August, 2019; originally announced August 2019.

Report number: SREP-19-29787

Journal ref: Sci Rep 10, 7829 (2020)

Speech Enhancement with Wide Residual Networks in Reverberant Environments

Authors: Jorge Llombart, Dayana Ribas, Antonio Miguel, Luis Vicente, Alfonso Ortega, Eduardo Lleida

Abstract: This paper proposes a speech enhancement method which exploits the high potential of residual connections in a Wide Residual Network architecture. This is supported on single dimensional convolutions computed alongside the time domain, which is a powerful approach to process contextually correlated representations through the temporal domain, such as speech feature sequences. We find the residual… ▽ More This paper proposes a speech enhancement method which exploits the high potential of residual connections in a Wide Residual Network architecture. This is supported on single dimensional convolutions computed alongside the time domain, which is a powerful approach to process contextually correlated representations through the temporal domain, such as speech feature sequences. We find the residual mechanism extremely useful for the enhancement task since the signal always has a linear shortcut and the non-linear path enhances it in several steps by adding or subtracting corrections. The enhancement capability of the proposal is assessed by objective quality metrics evaluated with simulated and real samples of reverberated speech signals. Results show that the proposal outperforms the state-of-the-art method called WPE, which is known to effectively reduce reverberation and greatly enhance the signal. The proposed model, trained with artificial synthesized reverberation data, was able to generalize to real room impulse responses for a variety of conditions (e.g. different room sizes, $RT_{60}$, near & far field). Furthermore, it achieves accuracy for real speech with reverberation from two different datasets. △ Less

Submitted 9 April, 2019; originally announced April 2019.

Comments: 5 pages, 4 figures. arXiv admin note: text overlap with arXiv:1901.00660, arXiv:1904.04511

Progressive Speech Enhancement with Residual Connections

Authors: Jorge Llombart, Dayana Ribas, Antonio Miguel, Luis Vicente, Alfonso Ortega, Eduardo Lleida

Abstract: This paper studies the Speech Enhancement based on Deep Neural Networks. The proposed architecture gradually follows the signal transformation during enhancement by means of a visualization probe at each network block. Alongside the process, the enhancement performance is visually inspected and evaluated in terms of regression cost. This progressive scheme is based on Residual Networks. During the… ▽ More This paper studies the Speech Enhancement based on Deep Neural Networks. The proposed architecture gradually follows the signal transformation during enhancement by means of a visualization probe at each network block. Alongside the process, the enhancement performance is visually inspected and evaluated in terms of regression cost. This progressive scheme is based on Residual Networks. During the process, we investigate a residual connection with a constant number of channels, including internal state between blocks, and adding progressive supervision. The insights provided by the interpretation of the network enhancement process leads us to design an improved architecture for the enhancement purpose. Following this strategy, we are able to obtain speech enhancement results beyond the state-of-the-art, achieving a favorable trade-off between dereverberation and the amount of spectral distortion. △ Less

Submitted 9 April, 2019; originally announced April 2019.

Comments: 5 pages, 5 figures

Optimization of the Area Under the ROC Curve using Neural Network Supervectors for Text-Dependent Speaker Verification

Authors: Victoria Mingote, Antonio Miguel, Alfonso Ortega, Eduardo Lleida

Abstract: This paper explores two techniques to improve the performance of text-dependent speaker verification systems based on deep neural networks. Firstly, we propose a general alignment mechanism to keep the temporal structure of each phrase and obtain a supervector with the speaker and phrase information, since both are relevant for a text-dependent verification. As we show, it is possible to use diffe… ▽ More This paper explores two techniques to improve the performance of text-dependent speaker verification systems based on deep neural networks. Firstly, we propose a general alignment mechanism to keep the temporal structure of each phrase and obtain a supervector with the speaker and phrase information, since both are relevant for a text-dependent verification. As we show, it is possible to use different alignment techniques to replace the global average pooling providing significant gains in performance. Moreover, we also present a novel back-end approach to train a neural network for detection tasks by optimizing the Area Under the Curve (AUC) as an alternative to the usual triplet loss function, so the system is end-to-end, with a cost function close to our desired measure of performance. As we can see in the experimental section, this approach improves the system performance, since our triplet neural network based on an approximation of the AUC (aAUC) learns how to discriminate between pairs of examples from the same identity and pairs of different identities. The different alignment techniques to produce supervectors in addition to the new back-end approach were tested on the RSR2015-Part I database for text-dependent speaker verification, providing competitive results compared to similar size networks using the global average pooling to extract supervectors and using a simple back-end or triplet loss training. △ Less

Submitted 30 April, 2019; v1 submitted 31 January, 2019; originally announced January 2019.

Disentangling and Learning Robust Representations with Natural Clustering

Authors: Javier Antoran, Antonio Miguel

Abstract: Learning representations that disentangle the underlying factors of variability in data is an intuitive way to achieve generalization in deep models. In this work, we address the scenario where generative factors present a multimodal distribution due to the existence of class distinction in the data. We propose N-VAE, a model which is capable of separating factors of variation which are exclusive… ▽ More Learning representations that disentangle the underlying factors of variability in data is an intuitive way to achieve generalization in deep models. In this work, we address the scenario where generative factors present a multimodal distribution due to the existence of class distinction in the data. We propose N-VAE, a model which is capable of separating factors of variation which are exclusive to certain classes from factors that are shared among classes. This model implements an explicitly compositional latent variable structure by defining a class-conditioned latent space and a shared latent space. We show its usefulness for detecting and disentangling class-dependent generative factors as well as its capacity to generate artificial samples which contain characteristics unseen in the training data. △ Less

Submitted 5 November, 2019; v1 submitted 27 January, 2019; originally announced January 2019.

Comments: Accepted at ICMLA 2019

Deep Speech Enhancement for Reverberated and Noisy Signals using Wide Residual Networks

Authors: Dayana Ribas, Jorge Llombart, Antonio Miguel, Luis Vicente

Abstract: This paper proposes a deep speech enhancement method which exploits the high potential of residual connections in a wide neural network architecture, a topology known as Wide Residual Network. This is supported on single dimensional convolutions computed alongside the time domain, which is a powerful approach to process contextually correlated representations through the temporal domain, such as s… ▽ More This paper proposes a deep speech enhancement method which exploits the high potential of residual connections in a wide neural network architecture, a topology known as Wide Residual Network. This is supported on single dimensional convolutions computed alongside the time domain, which is a powerful approach to process contextually correlated representations through the temporal domain, such as speech feature sequences. We find the residual mechanism extremely useful for the enhancement task since the signal always has a linear shortcut and the non-linear path enhances it in several steps by adding or subtracting corrections. The enhancement capacity of the proposal is assessed by objective quality metrics and the performance of a speech recognition system. This was evaluated in the framework of the REVERB Challenge dataset, including simulated and real samples of reverberated and noisy speech signals. Results showed that enhanced speech from the proposed method succeeded for both, the enhancement task with intelligibility purposes and the speech recognition system. The DNN model, trained with artificial synthesized reverberation data, was able to deal with far-field reverberated speech from real scenarios. Furthermore, the method was able to take advantage of the residual connection achieving to enhance signals with low noise level, which is usually a strong handicap of traditional enhancement methods. △ Less

Submitted 3 January, 2019; originally announced January 2019.

Tied Hidden Factors in Neural Networks for End-to-End Speaker Recognition

Authors: Antonio Miguel, Jorge Llombart, Alfonso Ortega, Eduardo Lleida

Abstract: In this paper we propose a method to model speaker and session variability and able to generate likelihood ratios using neural networks in an end-to-end phrase dependent speaker verification system. As in Joint Factor Analysis, the model uses tied hidden variables to model speaker and session variability and a MAP adaptation of some of the parameters of the model. In the training procedure our met… ▽ More In this paper we propose a method to model speaker and session variability and able to generate likelihood ratios using neural networks in an end-to-end phrase dependent speaker verification system. As in Joint Factor Analysis, the model uses tied hidden variables to model speaker and session variability and a MAP adaptation of some of the parameters of the model. In the training procedure our method jointly estimates the network parameters and the values of the speaker and channel hidden variables. This is done in a two-step backpropagation algorithm, first the network weights and factor loading matrices are updated and then the hidden variables, whose gradients are calculated by aggregating the corresponding speaker or session frames, since these hidden variables are tied. The last layer of the network is defined as a linear regression probabilistic model whose inputs are the previous layer outputs. This choice has the advantage that it produces likelihoods and additionally it can be adapted during the enrolment using MAP without the need of a gradient optimization. The decisions are made based on the ratio of the output likelihoods of two neural network models, speaker adapted and universal background model. The method was evaluated on the RSR2015 database. △ Less

Submitted 27 December, 2018; originally announced December 2018.

Journal ref: Proc. Interspeech 2017, 2819-2823

Differentiable Supervector Extraction for Encoding Speaker and Phrase Information in Text Dependent Speaker Verification

Authors: Victoria Mingote, Antonio Miguel, Alfonso Ortega, Eduardo Lleida

Abstract: In this paper, we propose a new differentiable neural network alignment mechanism for text-dependent speaker verification which uses alignment models to produce a supervector representation of an utterance. Unlike previous works with similar approaches, we do not extract the embedding of an utterance from the mean reduction of the temporal dimension. Our system replaces the mean by a phrase alignm… ▽ More In this paper, we propose a new differentiable neural network alignment mechanism for text-dependent speaker verification which uses alignment models to produce a supervector representation of an utterance. Unlike previous works with similar approaches, we do not extract the embedding of an utterance from the mean reduction of the temporal dimension. Our system replaces the mean by a phrase alignment model to keep the temporal structure of each phrase which is relevant in this application since the phonetic information is part of the identity in the verification task. Moreover, we can apply a convolutional neural network as front-end, and thanks to the alignment process being differentiable, we can train the whole network to produce a supervector for each utterance which will be discriminative with respect to the speaker and the phrase simultaneously. As we show, this choice has the advantage that the supervector encodes the phrase and speaker information providing good performance in text-dependent speaker verification tasks. In this work, the process of verification is performed using a basic similarity metric, due to simplicity, compared to other more elaborate models that are commonly used. The new model using alignment to produce supervectors was tested on the RSR2015-Part I database for text-dependent speaker verification, providing competitive results compared to similar size networks using the mean to extract embeddings. △ Less

Submitted 22 December, 2018; originally announced December 2018.

Comments: 5 pages, IberSPEECH 2018

gpuRIR: A Python Library for Room Impulse Response Simulation with GPU Acceleration

Authors: David Diaz-Guerra, Antonio Miguel, Jose R. Beltran

Abstract: The Image Source Method (ISM) is one of the most employed techniques to calculate acoustic Room Impulse Responses (RIRs), however, its computational complexity grows fast with the reverberation time of the room and its computation time can be prohibitive for some applications where a huge number of RIRs are needed. In this paper, we present a new implementation that dramatically improves the compu… ▽ More The Image Source Method (ISM) is one of the most employed techniques to calculate acoustic Room Impulse Responses (RIRs), however, its computational complexity grows fast with the reverberation time of the room and its computation time can be prohibitive for some applications where a huge number of RIRs are needed. In this paper, we present a new implementation that dramatically improves the computation speed of the ISM by using Graphic Processing Units (GPUs) to parallelize both the simulation of multiple RIRs and the computation of the images inside each RIR. Additional speedups were achieved by exploiting the mixed precision capabilities of the newer GPUs and by using lookup tables. We provide a Python library under GNU license that can be easily used without any knowledge about GPU programming and we show that it is about 100 times faster than other state of the art CPU libraries. It may become a powerful tool for many applications that need to perform a large number of acoustic simulations, such as training machine learning systems for audio signal processing, or for real-time room acoustics simulations for immersive multimedia systems, such as augmented or virtual reality. △ Less

Submitted 9 October, 2020; v1 submitted 26 October, 2018; originally announced October 2018.

Comments: This is a pre-print of an article published in Multimedia Tools and Applications (2020)

Analysis of the September $\varepsilon$-Perseid outburst in 2013

Authors: J. M. Madiedo, J. Zamorano, J. M. Trigo-Rodriguez, J. L. Ortiz, J. A. Docobo, J. Izquierdo, J. Lacruz, P. P. Campo, M. Andrade, S. Pastor, J. A. de los Reyes, F. Ocaña, A. Sanchez de Miguel, P. Pujols

Abstract: We analyze the outburst experienced by the September $\varepsilon$-Perseid meteor shower on 9 September 2013. As a result of our monitoring the atmospheric trajectory of 60 multi-station events observed over Spain was obtained and accurate orbital data were derived from them. On the basis of these orbits, we have tried to determine the likely parent body of this meteoroid stream by employing orbit… ▽ More We analyze the outburst experienced by the September $\varepsilon$-Perseid meteor shower on 9 September 2013. As a result of our monitoring the atmospheric trajectory of 60 multi-station events observed over Spain was obtained and accurate orbital data were derived from them. On the basis of these orbits, we have tried to determine the likely parent body of this meteoroid stream by employing orbital dissimilarity criteria. In addition, the emission spectra produced by two events belonging to this meteor shower were also recorded. The analysis of these spectra has provided information about the chemical nature of their progenitor meteoroids. We also present an estimation of the tensile strength for these particles. △ Less

Submitted 17 July, 2018; originally announced July 2018.

Comments: Accepted for publication in Monthly Notices of the Royal Astronomical Society on 16 July 2018

Estimating the relative contribution of streetlights, vehicles and residential lighting to the urban night sky brightness

Authors: Salvador Bará, Ángel Rodríguez-Arós, Marcos Pérez, Borja Tosar, Raul C. Lima, Alejandro Sánchez de Miguel, Jaime Zamorano

Abstract: Under stable atmospheric conditions, the zenithal brightness of the urban sky varies throughout the night following the time course of the anthropogenic emissions of light. Different types of artificial light sources (e.g. streetlights, residential, and vehicle lights) present specific time signatures, and this feature makes it possible to estimate the amount of sky brightness contributed by each… ▽ More Under stable atmospheric conditions, the zenithal brightness of the urban sky varies throughout the night following the time course of the anthropogenic emissions of light. Different types of artificial light sources (e.g. streetlights, residential, and vehicle lights) present specific time signatures, and this feature makes it possible to estimate the amount of sky brightness contributed by each one of them. Our approach is based on transforming the time representation of the zenithal sky brightness into a modal coefficients one, in terms of the time course signatures of the sources. The modal coefficients, and hence the absolute and relative contributions of each type of source, can be estimated from the measured brightness by means of linear least squares fits. A method for determining the time signatures is described, based on wide-field time-lapse photometry of the urban nightscape. Our preliminary results suggest that artificial light leaking out of the windows of residential buildings may account for a significant share of the time-varying part of the zenithal sky brightness, whilst the contribution of the vehicle lights seems to be significantly smaller. △ Less

Submitted 15 June, 2017; v1 submitted 14 June, 2017; originally announced June 2017.

Comments: ArXiv identifier added; acknowledgments updated

The history effect on bubble growth and dissolution. Part 2. Experiments and simulations of a spherical bubble attached to a horizontal flat plate

Authors: Pablo Peñas-López, Álvaro Moreno Soto Miguel, A. Parrales, Devaraj van der Meer, Detlef Lohse, Javier Rodríguez-Rodríguez

Abstract: The accurate description of the growth or dissolution dynamics of a soluble gas bubble in a super- or undersaturated solution requires taking into account a number of physical effects that contribute to the instantaneous mass transfer rate. One of these effects is the so-called history effect. It refers to the contribution of the local concentration boundary layer around the bubble that has develo… ▽ More The accurate description of the growth or dissolution dynamics of a soluble gas bubble in a super- or undersaturated solution requires taking into account a number of physical effects that contribute to the instantaneous mass transfer rate. One of these effects is the so-called history effect. It refers to the contribution of the local concentration boundary layer around the bubble that has developed from past mass transfer events between the bubble and liquid surroundings. In Part 1 of this work \citep{penas2016}, a theoretical treatment of this effect was given for a spherical, isolated bubble. Here, Part 2 provides an experimental and numerical study of the history effect regarding a spherical bubble attached to a horizontal flat plate and in the presence of gravity. The simulation technique developed in this paper is based on a streamfunction--vorticity formulation that may be applied to other flows where bubbles or drops exchange mass in the presence of a gravity field. Using this numerical tool, simulations are performed for the same conditions used in the experiments, in which the bubble is exposed to subsequent growth and dissolution stages, using step-wise variations in the ambient pressure. Besides proving the relevance of the history effect, the simulations highlight the importance that boundary-induced advection has to accurately describe bubble growth and shrinkage, i.e. the bubble radius evolution. In addition, natural convection has a significant influence that shows up in the velocity field even at short times, though, given the supersaturation conditions studied here, the bubble evolution is expected to be mainly diffusive. △ Less

Submitted 3 June, 2017; originally announced June 2017.

Journal ref: Journal of Fluid Mechanics, 820, 479-510 (2017)

Sky Quality Meter measurements in a colour changing world

Authors: A. Sánchez de Miguel, M. Aubé J. Zamorano, M. Kocifaj, J. Roby, C. Tapia

Abstract: The Sky Quality Meter (SQM) has become the most common device to track the evolution of the brightness of the sky from polluted regions to first class astronomical observatories. A vast database of SQM measurements already exists for many places in the world. Unfortunately, the SQM operates over a wide spectral band and its spectral response interacts with the sky's spectrum in a complex manner. T… ▽ More The Sky Quality Meter (SQM) has become the most common device to track the evolution of the brightness of the sky from polluted regions to first class astronomical observatories. A vast database of SQM measurements already exists for many places in the world. Unfortunately, the SQM operates over a wide spectral band and its spectral response interacts with the sky's spectrum in a complex manner. This is why the optical signals are difficult to interpret when the data are recorded in regions with different sources of artificial light. The brightness of the night sky is linked in a complex way to ground-based light emissions while taking into account atmospheric-induced optical distortion as well as spectral transformation from the underlying ground surfaces. While the spectral modulation of the sky's radiance has been recognized, it still remains poorly characterized and quantified. The impact of the SQM's spectral characteristics on the sky brightness measurements is here analysed for different light sources, including low and high pressure sodium lamps, PC-amber and white LEDs, metal halide, and mercury lamps. We show that a routine conversion of radiance to magnitude is difficult or rather impossible because the average wavelength depends on actual atmospheric and environment conditions, the spectrum of the source, and device specific properties. We correlate SQM readings with both the Johnson astronomical photometry bands and the human system of visual perception, assuming different lighting technologies. These findings have direct implications for the processing of SQM data and for its improvement and/or remediation. △ Less

Submitted 18 January, 2017; originally announced January 2017.

Comments: Accepted on MNRAS: ref. MN-16-1235-MJ.R2 Accepted 2017 January 17. Received 2017 January 12; in original form 2016 April 15

Low cost multi-purpose balloon-borne platform for wide-field imaging and video observation

Authors: Francisco Ocaña, Alejandro Sánchez de Miguel, Aitor Conde, Daedalus Team

Abstract: Atmosphere layers, especially the troposphere, hinder the astronomical observation. For more than 100 years astronomers have tried observing from balloons to avoid turbulence and extinction. New developments in cardsize computers, RF equipment and satellite navigation have democratised the access to the stratosphere. As a result of a ProAm collaboration with the Daedalus Team we have developed a l… ▽ More Atmosphere layers, especially the troposphere, hinder the astronomical observation. For more than 100 years astronomers have tried observing from balloons to avoid turbulence and extinction. New developments in cardsize computers, RF equipment and satellite navigation have democratised the access to the stratosphere. As a result of a ProAm collaboration with the Daedalus Team we have developed a low-cost multi-purpose platform with stratospheric balloons carrying up to 3 kg of scientific payload. The Daedalus Team is an amateur group that has been launching sounding probes since 2010. Since then the first two authors have provided scienti fic payloads for nighttime flights with the purpose of technology demonstration for astronomical observation. We have successfully observed meteor showers (Geminids 2012, Camelopardalis 2014, Quadrantids 2016 and Lyrids 2016) and city light pollution emission with image and video sensors covering the 400-1000nm range. △ Less

Submitted 26 September, 2016; originally announced September 2016.

Comments: 9 pages, 6 figures

Journal ref: Proceedings of the SPIE Astronomical Instrumentation and Telescopes conference, Ground-based and Airborne Telescopes VI (9906), 2016

Testing sky brightness models against radial dependency: a dense two dimensional survey around the city of Madrid, Spain

Authors: Jaime Zamorano, Alejandro Sánchez de Miguel, Francisco Ocaña, Berenice Pila-Diez, José Gómez Castaño, Sergio Pascual, Carlos Tapia, Jesús Gallego, Alberto Fernandez, Miguel Nievas

Abstract: We present a study of the night sky brightness around the extended metropolitan area of Madrid using Sky Quality Meter (SQM) photometers. The map is the first to cover the spatial distribution of the sky brightness in the center of the Iberian peninsula. These surveys are neccessary to test the light pollution models that predict night sky brightness as a function of the location and brightness of… ▽ More We present a study of the night sky brightness around the extended metropolitan area of Madrid using Sky Quality Meter (SQM) photometers. The map is the first to cover the spatial distribution of the sky brightness in the center of the Iberian peninsula. These surveys are neccessary to test the light pollution models that predict night sky brightness as a function of the location and brightness of the sources of light pollution and the scattering of light in the atmosphere. We describe the data-retrieval methodology, which includes an automated procedure to measure from a moving vehicle in order to speed up the data collection, providing a denser and wider survey than previous works with similar time frames. We compare the night sky brightness map to the nocturnal radiance measured from space by the DMSP satellite. We find that i) a single source model is not enough to explain the radial evolution of the night sky brightness, despite the predominance of Madrid in size and population, and ii) that the orography of the region should be taken into account when deriving geo-specific models from general first-principles models. We show the tight relationship between these two luminance measures. This finding sets up an alternative roadmap to extended studies over the globe that will not require the local deployment of photometers or trained personnel. △ Less

Submitted 21 December, 2015; originally announced December 2015.

Comments: Journal of Quantitative Spectroscopy & Radiative Transfer December 21, 2015

On the frequency of the superfireballs: more than 150 years of reports

Authors: Sandra Zamora, Fancisco Ocaña, Alejandro Sánchez de Miguel, Maruška Mole

Abstract: Superfireballs are rare phenomena for which the reports are scarce and the estimation of their abundance has a huge margin of uncertainty. As a citizen science project we have gathered >500 reports from newspapers in the 1850-2000 period. This database shows how some superfireball abundances are constant during the period, though the reference newspapers have changed in the last two centuries. We… ▽ More Superfireballs are rare phenomena for which the reports are scarce and the estimation of their abundance has a huge margin of uncertainty. As a citizen science project we have gathered >500 reports from newspapers in the 1850-2000 period. This database shows how some superfireball abundances are constant during the period, though the reference newspapers have changed in the last two centuries. We have tentatively related some fireball sources to well-known meteor showers (Perseids, Geminids and Leonids), while superfireball sources may be related to minor or unknown showers, probably of asteroidal origin. △ Less

Submitted 2 November, 2015; originally announced November 2015.

Comments: 4 pages, Proceedings of the IMC, Mistelbach, 2015

Orbits and emission spectra from the 2014 Camelopardalids

Authors: José M. Madiedo, Josep M. Trigo-Rodríguez, Jaime Zamorano, Jaime Izquierdo, Alejandro Sánchez de Miguel, Francisco Ocaña, José L. Ortiz, Francisco Espartero, Lorenzo G. Morillas, David Cardeñosa, Manuel Moreno-Ibáñez, Marta Urzáiz

Abstract: We have analyzed the meteor activity associated with meteoroids of fresh dust trails of Comet 209P/LINEAR, which produced an outburst of the Camelopardalid meteor shower (IAU code #451, CAM) in May 2014. With this aim, we have employed an array of high-sensitivity CCD video devices and spectrographs deployed at 10 meteor observing stations in Spain in the framework of the Spanish Meteor Network (S… ▽ More We have analyzed the meteor activity associated with meteoroids of fresh dust trails of Comet 209P/LINEAR, which produced an outburst of the Camelopardalid meteor shower (IAU code #451, CAM) in May 2014. With this aim, we have employed an array of high-sensitivity CCD video devices and spectrographs deployed at 10 meteor observing stations in Spain in the framework of the Spanish Meteor Network (SPMN). Additional meteoroid flux data were obtained by means of two forward-scatter radio systems. The observed peak zenithal hourly rate (ZHR) was much lower than expected, of around 20 meteors h-1. Despite of the small meteor flux in the optical range, we have obtained precise atmospheric trajectory, radiant and orbital information for 11 meteor and fireball events associated with this stream. The ablation behaviour and low tensile strength calculated for these particles reveal that Camelopardalid meteoroids are very fragile, mostly pristine aggregates with strength similar to that of the Orionids and the Leonids. The mineral grains seem to be glued together by a volatile phase. We also present and discuss two unique emission spectra produced by two Camelopardalid bright meteors. These suggest a non-chondritic nature for these particles, which exhibit Fe depletion in their composition. △ Less

Submitted 23 September, 2014; originally announced September 2014.

Comments: Accepted for publication in Monthly Notices of the Royal Astronomical Society on Sept. 22, 2014

Evolution of the energy consumed by street lighting in Spain estimated with DMSP-OLS data

Authors: Alejandro Sánchez de Miguel, Jaime Zamorano, José Gómez Castaño, Sergio Pascual

Abstract: We present the results of the analysis of satellite imagery to study light pollution in Spain. Both calibrated and non-calibrated DMSP-OLS images were used. We describe the method to scale the non-calibrated DMSP-OLS images which allows us to use differential photometry techniques in order to study the evolution of the light pollution. Population data and DMSP-OLS satellite calibrated images for t… ▽ More We present the results of the analysis of satellite imagery to study light pollution in Spain. Both calibrated and non-calibrated DMSP-OLS images were used. We describe the method to scale the non-calibrated DMSP-OLS images which allows us to use differential photometry techniques in order to study the evolution of the light pollution. Population data and DMSP-OLS satellite calibrated images for the year 2006 were compared to test the reliability of official statistics in public lighting consumption. We found a relationship between the population and the energy consumption which is valid for several regions. Finally the true evolution of the electricity consumption for street lighting in Spain from 1992 to 2010 was derived, it have been doubled in the last 18 years in most of the provinces. △ Less

Submitted 27 November, 2013; originally announced November 2013.

Comments: Accepted in JQSRT 26/11/2013 Light pollution special Issue

Report number: JQSRT-D-13-00221R1

Analysis of a superbolide from a Damocloid observed over Spain on July 13, 2012

Authors: Jose M. Madiedo, Josep M. Trigo-Rodriguez, Jaime Zamorano, Jose L. Ortiz, Alejandro Sanchez de Miguel, Francisco Ocaña, Jaime Izquierdo, Alberto J. Castro-Tirado, Nicolas Morales, David Galadi, Enrique de Guindos, Juan Lacruz, Faustino Organero, Leonor Ana-Hernandez, Fernando Fonseca, Mar Tapia, Felipe Gallego, Jesus Cabrera-Caño

Abstract: A superbolide with an estimated absolute magnitude of -20+-1 was seen on July 13, 2012 over the center and south of Spain. This extraordinary event, which was witnessed by numerous casual observers, was recorded in the framework of the continuous fireball monitoring and meteor spectroscopy campaign performed by the SPanish Meteor Network (SPMN). Thus, because of optimal weather conditions, the bol… ▽ More A superbolide with an estimated absolute magnitude of -20+-1 was seen on July 13, 2012 over the center and south of Spain. This extraordinary event, which was witnessed by numerous casual observers, was recorded in the framework of the continuous fireball monitoring and meteor spectroscopy campaign performed by the SPanish Meteor Network (SPMN). Thus, because of optimal weather conditions, the bolide was imaged from ten meteor observing stations. Here we present the analysis of this magnificent event, which is the brightest fireball ever recorded by our team. The atmospheric trajectory of the bolide and the orbit in the Solar System of the parent meteoroid were obtained. The emission spectrum produced during the ablation of this particle is also discussed. We found that the meteoroid, which was following a Halley Type Comet orbit, was depleted in Na and had a tensile strength one order of magnitude higher than that corresponding to typical cometary materials. By means of orbital analysis tools we have investigated the likely parent body of this particle and the results suggest that the progenitor is a damocloid. The impact area of the hypothetical remnants of the meteoroid is also given and a search for meteorites was performed, but none were found. △ Less

Submitted 1 October, 2013; originally announced October 2013.

Comments: Paper accepted for publication in Monthly Notices of the Royal Astronomical Society (MNRAS) on Sept. 26, 2013

The 2011 October Draconids outburst. I. Orbital elements, meteoroid fluxes and 21P/Giacobini-Zinner delivered mass to Earth

Authors: Josep M. Trigo-Rodríguez, José M. Madiedo, I. P. Williams, Joan Dergham, Jordi Cortés, Alberto J. Castro-Tirado, José L. Ortiz, Jaime Zamorano, Francisco Ocaña, Jaime Izquierdo, Alejandro Sánchez de Miguel, Jacinto Alonso-Azcárate, Diego Rodríguez, Mar Tapia, Pep Pujols, Juan Lacruz, Francesc Pruneda, Armand Oliva, Juan Pastor Erades, Antonio Francisco Marín

Abstract: On October 8th, 2011 the Earth crossed the dust trails left by comet 21P/Giacobini-Zinner during its XIX and XX century perihelion approaches with the comet being close to perihelion. The geometric circumstances of that encounter were thus favorable to produce a meteor storm, but the trails were much older than in the 1933 and 1946 historical encounters. As a consequence the 2011 October Draconid… ▽ More On October 8th, 2011 the Earth crossed the dust trails left by comet 21P/Giacobini-Zinner during its XIX and XX century perihelion approaches with the comet being close to perihelion. The geometric circumstances of that encounter were thus favorable to produce a meteor storm, but the trails were much older than in the 1933 and 1946 historical encounters. As a consequence the 2011 October Draconid display exhibited several activity peaks with Zenithal Hourly Rates of about 400 meteors per hour. In fact, if the display had been not forecasted, it could have passed almost unnoticed as was strongly attenuated for visual observers due to the Moon. This suggests that most meteor storms of a similar nature could have passed historically unnoticed under unfavorable weather and Moon observing conditions. The possibility of obtaining information on the physical properties of cometary meteoroids penetrating the atmosphere under low-geocentric velocity encounter circumstances motivated us to set up a special observing campaign. Added to the Spanish Fireball Network wide-field all-sky and CCD video monitoring, other high-sensitivity 1/2" black and white CCD video cameras were attached to modified medium-field lenses for obtaining high resolution orbital information. The trajectory, radiant, and orbital data of 16 October Draconid meteors observed at multiple stations are presented. The results show that the meteors appeared from a geocentric radiant located at R.A.=263.0+-0.4 deg. and Dec.=+55.3+-0.3 deg. that is in close agreement with the radiant predicted for the 1873-1894 and the 1900 dust trails. The estimated mass of material from 21P/Giacobini-Zinner delivered to Earth during the six-hours outburst was around 950+-150 kg. △ Less

Submitted 29 April, 2013; originally announced April 2013.

Comments: Manuscript in press in Monthly Notices of the Royal Astronomical Society, submitted to MNRAS on November 16th, 2012 Accepted for publication in MNRAS on April 28th, 2013 Manuscript Pages: 21 Tables: 8 Figures: 4 Manuscript associated: "The 2011 October Draconids outburst. II. Meteoroid chemical abundances from fireball spectroscopy" by J.M. Madiedo is also in press in the same journal

Report number: MNRAS reference# MN-12-2858-MJ.R2 MSC Class: 70F15 ACM Class: F.2.2; I.2.7

Journal ref: Trigo-Rodríguez J.M. et al. (2013), Monthly Notices of the Royal Astronomical Society

The crystal structure of cold compressed graphite

Authors: Maximilian Amsler, José A. Flores-Livas, Lauri Lehtovaara, Felix Balima, S. Alireza Ghasemi, Denis Machon, Stéphane Pailhès, Alexander Willand, Damien Caliste, Silvana Botti, Alfonso San Miguel, Stefan Goedecker, Miguel A. L. Marques

Abstract: Through a systematic structural search we found an allotrope of carbon with Cmmm symmetry which we predict to be more stable than graphite for pressures above 10 GPa. This material, which we refer to as Z-carbon, is formed by pure sp3 bonds and is the only carbon allotrope which provides an excellent match to unexplained features in experimental X-ray diffraction and Raman spectra of graphite unde… ▽ More Through a systematic structural search we found an allotrope of carbon with Cmmm symmetry which we predict to be more stable than graphite for pressures above 10 GPa. This material, which we refer to as Z-carbon, is formed by pure sp3 bonds and is the only carbon allotrope which provides an excellent match to unexplained features in experimental X-ray diffraction and Raman spectra of graphite under pressure. The transition from graphite to Z-carbon can occur through simple sliding and buckling of graphene sheets. Our calculations predict that Z-carbon is a transparent wide band gap semiconductor with a hardness comparable to diamond. △ Less

Submitted 6 September, 2011; originally announced September 2011.

Comments: 4 pages, 5 figures

Journal ref: Phys. Rev. Lett. 108, 117004 (2012)

Ab initio study of charge do** effect on 1D polymerization of C60

Authors: Roberta Poloni, Alfonso San Miguel, Maria Victoria Fernandez-Serra

Abstract: We study the interplay between charge do** and intermolecular distance in the polymerization of C60 fullerene chains by means of density functional theory (DFT)-based first principle calculations. The potential energy surface analysis shows that both the equilibrium intermolecular distance of the unpolymerized system and the polymerization energy barrier are inversely proportional to the electro… ▽ More We study the interplay between charge do** and intermolecular distance in the polymerization of C60 fullerene chains by means of density functional theory (DFT)-based first principle calculations. The potential energy surface analysis shows that both the equilibrium intermolecular distance of the unpolymerized system and the polymerization energy barrier are inversely proportional to the electron do** of the system. We analyze the origin of this charge-induced polymerization effect by studying the behavior of the system's wavefunctions around the Fermi level and the structural modifications of the molecules as a function of two variables: the distance between the centers of the molecules and the number of electrons added to the system. △ Less

Submitted 16 May, 2011; originally announced May 2011.

Journal ref: J. Phys.: Cond. Matter 24 (2012) 095501