Search

In this study, we investigate if noise-augmented training can concurrently improve adversarial robustness in automatic speech recognition (ASR) systems. We conduct a comparative analysis of the adversarial robustness of four different state-of-the-art ASR architectures, where each of the ASR architectures is trained under three different augmentation conditions: one subject to background noise, speed variations, and reverberations, another subject to speed variations only, and a third without any form of data augmentation. The results demonstrate that noise augmentation not only improves model performance on noisy speech but also the model's robustness to adversarial attacks.

This paper presents a natural language processing (NLP) approach to the problem of thoroughly comprehending song lyrics, with particular attention on genre classification, view-based success prediction, and approximate release year. Our tests provide promising results with 65\% accuracy in genre classification and 79\% accuracy in success prediction, leveraging a DistilBERT model for genre classification and BERT embeddings for release year prediction. Support Vector Machines outperformed other models in predicting the release year, achieving the lowest root mean squared error (RMSE) of 14.18. Our study offers insights that have the potential to revolutionize our relationship with music by addressing the shortcomings of current approaches in properly understanding the emotional intricacies of song lyrics.

While computational thinking arises as an essential skill worldwide, formal primary and secondary education in Latin America rarely incorporates mechanisms to develop it in their curricula. The extent to which students in the region acquire computational thinking skills remains largely unknown. To start addressing this void, this article presents findings from a cross-sectional study that characterizes the computational thinking abilities of incoming students at a Chilean university with a strong emphasis on STEM disciplines. Based on more than 500 responses, this study provides evidence of significant inequalities in computational thinking across gender, type of school (private or no), and prior programming knowledge. The discussion offers insights into how these disparities relate to contextual factors of the country, such as a highly socio-economically segregated educational system, public policies focused mainly on technology access, and heavy reliance on voluntary initiatives, to develop computational thinking. The findings can enlighten upcoming research endeavors and formulate strategies to create a more equitable field for students entering STEM degrees in nations facing similar circumstances., Comment: Pre-print. Final version published in: IEEE Transactions on Education

We investigate the task of identifying situations of distracted driving through analysis of in-car videos. To tackle this challenge we introduce a multi-task video transformer that predicts both distracted actions and driver pose. Leveraging VideoMAEv2, a large pre-trained architecture, our approach incorporates semantic information from human keypoint locations to enhance action recognition and decrease computational overhead by minimizing the number of spatio-temporal tokens. By guiding token selection with pose and class information, we notably reduce the model's computational requirements while preserving the baseline accuracy. Our model surpasses existing state-of-the art results in driver action recognition while exhibiting superior efficiency compared to current video transformer-based approaches.

One use of Autonomous Underwater Vehicles (AUVs) is the monitoring of habitats associated with threatened, endangered and protected marine species, such as the handfish of Tasmania, Australia. Seafloor imagery collected by AUVs can be used to identify individuals within their broader habitat context, but the sheer volume of imagery collected can overwhelm efforts to locate rare or cryptic individuals. Machine learning models can be used to identify the presence of a particular species in images using a trained object detector, but the lack of training examples reduces detection performance, particularly for rare species that may only have a small number of examples in the wild. In this paper, inspired by recent work in few-shot learning, images and annotations of common marine species are exploited to enhance the ability of the detector to identify rare and cryptic species. Annotated images of six common marine species are used in two ways. Firstly, the common species are used in a pre-training step to allow the backbone to create rich features for marine species. Secondly, a copy-paste operation is used with the common species images to augment the training data. While annotations for more common marine species are available in public datasets, they are often in point format, which is unsuitable for training an object detector. A popular semantic segmentation model efficiently generates bounding box annotations for training from the available point annotations. Our proposed framework is applied to AUV images of handfish, increasing average precision by up to 48\% compared to baseline object detection training. This approach can be applied to other objects with low numbers of annotations and promises to increase the ability to actively monitor threatened, endangered and protected species., Comment: 7 pages, 5 figures. Submitted to the 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2024)

CHEOPS is a space telescope specifically designed to monitor transiting exoplanets orbiting bright stars. In September 2023, CHEOPS completed its nominal mission and remains in excellent operational conditions. The mission has been extended until the end of 2026. Scientific and instrumental data have been collected throughout in-orbit commissioning and nominal operations, enabling a comprehensive analysis of the mission's performance. In this article, we present the results of this analysis with a twofold goal. First, we aim to inform the scientific community about the present status of the mission and what can be expected as the instrument ages. Secondly, we intend for this publication to serve as a legacy document for future missions, providing insights and lessons learned from the successful operation of CHEOPS. To evaluate the instrument performance in flight, we developed a comprehensive monitoring and characterisation programme. It consists of dedicated observations that allow us to characterise the instrument's response. In addition to the standard collection of nominal science and housekeeping data, these observations provide input for detecting, modelling, and correcting instrument systematics, discovering and addressing anomalies, and comparing the instrument's actual performance with expectations. The precision of the CHEOPS measurements has enabled the mission objectives to be met and exceeded. Careful modelling of the instrumental systematics allows the data quality to be significantly improved during the light curve analysis phase, resulting in more precise scientific measurements. CHEOPS is compliant with the driving scientific requirements of the mission. Although visible, the ageing of the instrument has not affected the mission's performance., Comment: Accepted for publication in Astronomy and Astrophysics

Samples of the monolithic silicon pixel ASIC prototype produced in 2022 within the framework of the Horizon 2020 MONOLITH ERC Advanced project were irradiated with 70 MeV protons up to a fluence of 1 x 1016 neq/cm2, and then tested using a beam of 120 GeV/c pions. The ASIC contains a matrix of 100 \mu m pitch hexagonal pixels, readout out by low noise and very fast frontend electronics produced in a 130 nm SiGe BiCMOS technology process. The dependence on the proton fluence of the efficiency and the time resolution of this prototype was measured with the frontend electronics operated at a power density between 0.13 and 0.9 W/cm2. The testbeam data show that the detection efficiency of 99.96% measured at sensor bias voltage of 200 V before irradiation becomes 96.2% after a fluence of 1 x 1016 neq/cm2. An increase of the sensor bias voltage to 300 V provides an efficiency to 99.7% at that proton fluence. The timing resolution of 20 ps measured before irradiation rises for a proton fluence of 1 x 1016 neq/cm2 to 53 and 45 ps at HV = 200 and 300 V, respectively.

Neural topic models can successfully find coherent and diverse topics in textual data. However, they are limited in dealing with multimodal datasets (e.g., images and text). This paper presents the first systematic and comprehensive evaluation of multimodal topic modeling of documents containing both text and images. In the process, we propose two novel topic modeling solutions and two novel evaluation metrics. Overall, our evaluation on an unprecedented rich and diverse collection of datasets indicates that both of our models generate coherent and diverse topics. Nevertheless, the extent to which one method outperforms the other depends on the metrics and dataset combinations, which suggests further exploration of hybrid solutions in the future. Notably, our succinct human evaluation aligns with the outcomes determined by our proposed metrics. This alignment not only reinforces the credibility of our metrics but also highlights the potential for their application in guiding future multimodal topic modeling endeavors., Comment: Camera-Ready for LREC-COLING 2024 (Long Paper)

This paper presents the first results of the study of high-energy electron and muon neutrino charged-current interactions in the FASER$\nu$ emulsion/tungsten detector of the FASER experiment at the LHC. A subset of the FASER$\nu$ volume, which corresponds to a target mass of 128.6~kg, was exposed to neutrinos from the LHC $pp$ collisions with a centre-of-mass energy of 13.6~TeV and an integrated luminosity of 9.5 fb$^{-1}$. Applying stringent selections requiring electrons with reconstructed energy above 200~GeV, four electron neutrino interaction candidate events are observed with an expected background of $0.025^{+0.015}_{-0.010}$, leading to a statistical significance of 5.2$\sigma$. This is the first direct observation of electron neutrino interactions at a particle collider. Eight muon neutrino interaction candidate events are also detected, with an expected background of $0.22^{+0.09}_{-0.07}$, leading to a statistical significance of 5.7$\sigma$. The signal events include neutrinos with energies in the TeV range, the highest-energy electron and muon neutrinos ever detected from an artificial source. The energy-independent part of the interaction cross section per nucleon is measured over an energy range of 560--1740 GeV (520--1760 GeV) for $\nu_e$ ($\nu_{\mu}$) to be $(1.2_{-0.7}^{+0.8}) \times 10^{-38}~\mathrm{cm}^{2}\,\mathrm{GeV}^{-1}$ ($(0.5\pm0.2) \times 10^{-38}~\mathrm{cm}^{2}\,\mathrm{GeV}^{-1}$), consistent with Standard Model predictions. These are the first measurements of neutrino interaction cross sections in those energy ranges.

Fine-tuning the pre-trained model with active learning holds promise for reducing annotation costs. However, this combination introduces significant computational costs, particularly with the growing scale of pre-trained models. Recent research has proposed proxy-based active learning, which pre-computes features to reduce computational costs. Yet, this approach often incurs a significant loss in active learning performance, which may even outweigh the computational cost savings. In this paper, we argue the performance drop stems not only from pre-computed features' inability to distinguish between categories of labeled samples, resulting in the selection of redundant samples but also from the tendency to compromise valuable pre-trained information when fine-tuning with samples selected through the proxy model. To address this issue, we propose a novel method called aligned selection via proxy to update pre-computed features while selecting a proper training method to inherit valuable pre-training information. Extensive experiments validate that our method significantly improves the total cost of efficient active learning while maintaining computational efficiency.

The Forward Search Experiment (FASER) at CERN's Large Hadron Collider (LHC) has recently directly detected the first collider neutrinos. Neutrinos play an important role in all FASER analyses, either as signal or background, and it is therefore essential to understand the neutrino event rates. In this study, we update previous simulations and present prescriptions for theoretical predictions of neutrino fluxes and cross sections, together with their associated uncertainties. With these results, we discuss the potential for possible measurements that could be carried out in the coming years with the FASER neutrino data to be collected in LHC Run 3 and Run 4., Comment: 19 pages, 10 figures

The research interest on location-based services has increased during the last years ever since 3D centimetre accuracy inside intelligent environments could be confronted with. This work proposes an indoor local positioning system based on LED lighting, transmitted from a set of beacons to a receiver.The receiver is based on a quadrant photodiode angular diversity aperture (QADA) plus an aperture placed over it.This configuration can be modelled as a perspective camera, where the image position of the transmitters can be used to recover the receiver's 3D pose. This process is known as the perspective-n-point (PnP) problem, which is well known in computer vision and photogrammetry. This work investigates the use of different state-of-the-art PnP algorithms to localize the receiver in a large space based on four co-planar transmitters and with a distance from transmitters to receiver of 3.4 m. Encoding techniques are used to permit the simultaneous emission of all the transmitted signals and their processing in the receiver. In addition, correlation techniques are used to determine the image points projected from each emitter on the QADA. This work uses Monte Carlo simulations to characterize the absolute errors for a grid of test points under noisy measurements, as well as the robustness of the system when varying the 3D location of one transmitter. The IPPE algorithm obtained the best performance in this configuration. The proposal has also been experimentally evaluated in a real setup. The estimation of the receiver's position at three points using the IPPE algorithm achieves average absolute errors of 4.33cm, 3.51cm and 28.90cm in the coordinates x, y and z, respectively. These positioning results are in line with those obtained in previous work using triangulation techniques but with the addition that the complete pose of the receiver is obtained in this proposal.

The time resolution of the second monolithic silicon pixel prototype produced for the MONOLITH H2020 ERC Advanced project was studied using a femtosecond laser. The ASIC contains a matrix of hexagonal pixels with 100 {\mu}m pitch, readout by low-noise and very fast SiGe HBT frontend electronics. Silicon wafers with 50 {\mu}m thick epilayer with a resistivity of 350 {\Omega}cm were used to produce a fully depleted sensor. At the highest frontend power density tested of 2.7 W/cm2, the time resolution with the femtosecond laser pulses was found to be 45 ps for signals generated by 1200 electrons, and 3 ps in the case of 11k electrons, which corresponds approximately to 0.4 and 3.5 times the most probable value of the charge generated by a minimum-ionizing particle. The results were compared with testbeam data taken with the same prototype to evaluate the time jitter produced by the fluctuations of the charge collection., Comment: Submitted to JINST

The sky-averaged cosmological 21 cm signal can improve our understanding of the evolution of the early Universe from the Dark Age to the end of the Epoch of Reionization. Although the EDGES experiment reported an absorption profile of this signal, there have been concerns about the plausibility of these results, motivating independent validation experiments. One of these initiatives is the Mapper of the IGM Spin Temperature (MIST), which is planned to be deployed at different remote locations around the world. One of its key features is that it seeks to comprehensively compensate for systematic uncertainties through detailed modeling and characterization of its different instrumental subsystems, particularly its antenna. Here we propose a novel optimizing scheme which can be used to design an antenna applied to MIST, improving bandwidth, return loss, and beam chromaticity. This new procedure combines the Particle Swarm Optimization (PSO) algorithm with a commercial electromagnetic simulation software (HFSS). We improved the performance of two antenna models: a rectangular blade antenna, similar to the one used in the EDGES experiment, and a trapezoidal bow-tie antenna. Although the performance of both antennas improved after applying our optimization method, we found that our bow-tie model outperforms the blade antenna by achieving lower reflection losses and beam chromaticity in the entire band of interest. To further validate the optimization process, we also built and characterized 1:20 scale models of both antenna types showing an excellent agreement with our simulations., Comment: 16 pages, published in the Journal of Astronomical Instrumentation

Research on developing deep learning techniques for autonomous spacecraft relative navigation challenges is continuously growing in recent years. Adopting those techniques offers enhanced performance. However, such approaches also introduce heightened apprehensions regarding the trustability and security of such deep learning methods through their susceptibility to adversarial attacks. In this work, we propose a novel approach for adversarial attack detection for deep neural network-based relative pose estimation schemes based on the explainability concept. We develop for an orbital rendezvous scenario an innovative relative pose estimation technique adopting our proposed Convolutional Neural Network (CNN), which takes an image from the chaser's onboard camera and outputs accurately the target's relative position and rotation. We perturb seamlessly the input images using adversarial attacks that are generated by the Fast Gradient Sign Method (FGSM). The adversarial attack detector is then built based on a Long Short Term Memory (LSTM) network which takes the explainability measure namely SHapley Value from the CNN-based pose estimator and flags the detection of adversarial attacks when acting. Simulation results show that the proposed adversarial attack detector achieves a detection accuracy of 99.21%. Both the deep relative pose estimator and adversarial attack detector are then tested on real data captured from our laboratory-designed setup. The experimental results from our laboratory-designed setup demonstrate that the proposed adversarial attack detector achieves an average detection accuracy of 96.29%.