One of the most rapidly advancing areas of deep learning research aims at creating models that learn to disentangle the latent factors of variation from a data distribution. However, modeling joint probability mass functions is usually prohibitive, which motivates the use of conditional models assuming that some information is given as input. In the domain of numerical cognition, deep learning architectures have successfully demonstrated that approximate numerosity representations can emerge in multi-layer networks that build latent representations of a set of images with a varying number of items. However, existing models have focused on tasks requiring to conditionally estimate numerosity information from a given image. Here we focus on a set of much more challenging tasks, which require to conditionally generate synthetic images containing a given number of items. We show that attention-based architectures operating at the pixel level can learn to produce well-formed images approximately containing a specific number of items, even when the target numerosity was not present in the training distribution.

Healthcare professionals and resource planners can use healthcare delivery process mining to ensure the optimal utilisation of scarce healthcare resources when developing policies. Within hospitals, patients' Length of Stay (LOS) and volume of admitted patients, in terms of number and characteristics (age, gender, and social deter-minants), are significant factors determining daily resource requirements. In this study, we used Coxian phase-type Distribution (C-PHD) based Phase-Type Survival (PTS) trees for analysing how covariates such as admission date, gender, age, district, and admissions source influence the admission rate and LOS distribution. PTS trees. This study used a two-year data set (2011-2012) of patients admitted to the Emergency Department at Mater Dei Hospital to generate models and an independent one-year data set (2013) of patients admitted to the Emergency Department at Mater Dei Hospital to evaluate. The PTS tree effectively clusters patients based on their LOS, considering the prognostic significance of different covariates related to patients' characteristics. Charac-terising these covariates provided meaningful results about LOS. Similarly, the PTS tree was used to effectively cluster patients based on the admission rate, considering the prognostic significance of these covariates.

Constructing an innovation system centered on new-quality productivity is key to achieving China's technological independence and high-quality development, especially against the backdrop of rapid advancements in artificial intelligence. This article utilizes provincial panel data from China spanning 2010 to 2022 and applies kernel density estimation and the Moran's I index to study the spatiotemporal evolution of artificial intelligence and new-quality productivity. The findings indicate a clustering trend in the development of both factors across regions. Causal empirical analysis using a two-way fixed-effects model reveals that artificial intelligence positively influences new-quality productivity, with conclusions validated by robustness and endogeneity tests.

Given the high number of vehicle-crash victims, it has been established as a priority to reduce this figure in the transportation sector. For this reason, many of the recent researches are focused on including control systems in existing vehicles, to improve their stability, comfort and handling. These systems need to know in every moment the behavior of the vehicle (state variables), among others, when the different maneuvers are performed, to actuate by means of the systems in the vehicle (brakes, steering, suspension) and, in this way, to achieve a good behavior. The main problem arises from the lack of ability to directly capture several required dynamic vehicle variables, such as roll angle, from low-cost sensors. Previous studies demonstrate that low-cost sensors can provide data in real-time with the required precision and reliability. Even more, other research works indicate that neural networks are efficient mechanisms to estimate roll angle. Nevertheless, it is necessary to assess that the fusion of data coming from low-cost devices and estimations provided by neural networks can fulfill the reliability and appropriateness requirements for using these technologies to improve overall safety in production vehicles. Because of the increasing of computing power, the reduction of consumption and electric devices size, along with the high variety of communication technologies and networking protocols using Internet have yield to Internet of Things (IoT) development. In order to address this issue, this study has two main goals: 1) Determine the appropriateness and performance of neural networks embedded in low-cost sensors kits to estimate roll angle required to evaluate rollover risk situations. 2) Compare the low-cost control unit devices (Intel Edison and Raspberry Pi 3 Model B), to provide the roll angle estimation with this artificial neural network-based approach. To fulfil these objectives an experimental environment has been set up composed of a van with two set of low-cost kits, one including a Raspberry Pi 3 Model B, low cost Inertial Measurement Unit (BNO055 - 37€) and GPS (Mtk3339 - 53€) and the other having an Intel Edison System on Chip linked to a SparkFun 9 Degrees of Freedom module. This experimental environment will be tested in different maneuvers for comparison purposes. Neural networks embedded in low-cost sensor kits provide roll angle estimations very approximated to real values. Even more, Intel Edison and Raspberry Pi 3 Model B have enough computing capabilities to successfully run roll angle estimation based on neural networks to determine rollover risks situation fulfilling real-time operation restrictions stated for this problem.

Despite the increasing role of machine learning in various fields, very few works considered artificial intelligence for frequency estimation (FE). This work presents a comprehensive analysis of deep-learning (DL) approach for frequency estimation of single-tones. It is shown that DL network with two layers having a few nodes can estimate frequency more accurately than well-known classical techniques. The study is comprehensive, filling gaps of existing works, where it analyzes error under different signal-to-noise ratios, numbers of nodes, and numbers of input samples; also, under missing SNR information. It is found that DL-based FE is not significantly affected by SNR bias or number of nodes. DL-based approach can work properly using minimal number of input nodes N at which classical methods fail. It is possible for DL to use as little as two layers with two or three nodes each, with complexity of O{N} versus O{Nlog2 (N)} for DFT-based FE, noting that less N is required for DL. Hence, DL can significantly reduce FE complexity, memory, cost, and power consumption, making DL-based FE attractive for resource-limited systems like some IoT sensor applications. Also, reduced complexity opens the door for hardware-efficient implementation using short word-length (SWL) or time-efficient software-defined radio (SDR) communications.

Conclusions: Our findings suggest that the NMS-E represents a valuable preoperative diagnostic tool for endometriosis, effectively correlating with the disease's severity and surgical outcomes. Incorporating the NMS-E into clinical practice could significantly enhance the management of endometriosis by addressing current diagnostic limitations and guiding surgical planning.

Estimation of core and surface temperature is one of the crucial functionalities of the lithium-ion Battery Management System (BMS) towards providing effective thermal management, fault detection and operational safety. While, it is impractical to measure core temperature using physical sensors, implementing a complex estimation strategy in on-board low-cost BMS is challenging due to high computational cost and the cost of implementation. Typically, a temperature estimation scheme consists of a heat generation model and a heat transfer model. Several researchers have already proposed ranges of thermal models having different levels of accuracy and complexity. Broadly, there are first-order and second-order heat capacitor-resistor-based thermal models of lithium-ion batteries (LIBs) for core and surface temperature estimation. This paper deals with a detailed comparative study between these two models using extensive laboratory test data and simulation study to access suitability in online prediction and onboard BMS. The aim is to guide whether it’s worth investing towards developing a second-order model instead of a first-order model with respect to prediction accuracy considering modelling complexity, experiments required and the computational cost. Both the thermal models along with the parameter estimation scheme are modelled and simulated using MATLAB/Simulink environment. Models are validated using laboratory test data of a cylindrical 18650 LIB cell. Further, a Kalman Filter with appropriate process and measurement noise levels are used to estimate the core temperature in terms of measured surface and ambient temperatures. Results from the first-order model and second-order models are analyzed for comparison purposes.

Synthetic Aperture RADAR (SAR) is a radar imaging technique in which the relative motion of the sensor is used to synthesize a very long antenna and obtain high spatial resolution. Standard SAR raw data processing techniques assume uniform motion of the satellite (or aerial vehicle) and a fixed antenna beam pointing sideway orthogonally to the motion path, assumed rectilinear. Despite SAR data processing is a well established imaging technology that has become fundamental in several fields and applications, in this paper a novel approach has been used to exploit coherent illumination, demonstrating the possibility of extracting a large part of the ancillary data information from the raw data itself, to be used in the focusing procedure. In this work an effort has been carried out to try to focus the raw SAR complex data matrix without the knowledge of anyof the parameters needed in standard focusing procedures as Range Doppler (RD) algorithm, ω - K algorithm and Chirp Scaling (CS) algorithm. All the literature references regarding the algorithms needed to obtain a precise image from raw data use such parameters that refer both to the SAR system acquisition geometry and its radiometric specific parameters. In [12], authors introduced a preliminary work dealing with this problem and able to obtain, in the presence of a strong point scatterer in the observed scene, good quality images, if compared to the standard processing techniques. In this work the proposed technique is described and performances parameters are extracted to compare the proposed approach to RD.

Advanced methods of treatment are needed to fight the threats of virus-transmitted deseases and pandemics. Often they are based on an improved biophysical understanding of virus replication strategies and processes in their host cells. For instance, an essential component of the replication of the Hepatitis C virus (HCV) proceeds under the influence of non-structural HCV proteins (NSPs) that are anchored to the endoplasmatic reticulum (ER), such as the NS5a protein. The diffusion of NSPs has been studied by in-vitro fluorescence recovery after photobleaching (FRAP) experiments. The diffusive evolution of the concentration field of NSPs on the ER can be described by means of surface partial differential equations (sufPDE). Knodel et al, Viruses 2018, 10, 28 estimated the diffusion coefficient of the NS5a protein by minimizing the discrepancy of an extended set of sufPDE simulations and experimental FRAP time series data. Here, we provide a scaling analysis of the sufPDEs that describe the diffusive evolution of the concentration field of NSPs on the ER. This analysis provides an estimate of the diffusion coefficient that is based only on the ratio of the membrane surface area in the FRAP region and its contour length. The quality of this estimate is explored by comparison to numerical solutions of the sufPDE for a flat geometry and for ten different 3D embedded 2D ER grids that are derived from fluorescence z-stack data of the ER. Finally, we apply the new data analysis to the experimental FRAP time-series data analyzed in our previous paper, and we discuss the opportunities of the new approach.

The human ankle is a complex joint, most commonly represented as talocrural and subtalar axes. It is difficult to locate and take in vivo measurements of the ankle joint. There are no instruments for patients lying on a bed or the floor; that can be used in outdoor or remote sites. We have developed a "Turmell-meter" to address these issues. We started with the study of ankle anatomy and anthropometry, then we used the product of exponentials’ formula to visualize the movements. Furthermore, we built a prototype using human proportions and statistics. For pose estimation, we used a trilateration method by applying tetrahedral geometry. Additionally, we computed the axis direction by fitting 3D circles, plotting the manifold and chart as an ankle joint model. We presented the results of simulations, a prototype comprising 45 parts, specifically designed draw-wire sensors, and electronics. Finally, we tested the device by capturing positions and fitting them into the bi-axial ankle model as a Riemannian manifold. The Turmell-meter is intended to be a hardware platform for human ankle joint axis estimation, it is adjustable and has an easy setup. The proposed model has the properties of a chart in a geometric manifold, we provided the details

Recent improvements in deepfake creation have made deepfake videos more realistic. Moreover, open-source software has made deepfake creation more accessible, which reduces the barrier to entry for deepfake creation. This could pose a threat to the people’s privacy. There is a potential danger if the deepfake creation techniques are used by people with an ulterior motive to produce deepfake videos of world leaders to disrupt the order of countries and the world. Therefore, research into the automatic detection of deepfaked media is essential for public security. In this work, we propose a deepfake detection method using upper body language analysis. Specifically, a many-to-one LSTM network was designed and trained as a classification model for deepfake detection. Different models were trained by varying the hyperparameters to build a final model with benchmark accuracy. We achieved 94.39% accuracy on the deepfake test set. The experimental results showed that upper body language can effectively detect deepfakes.

The building sector has a big potential to reduce the material resource demand needed for building construction and therefore, greenhouse gas (GHG) emissions. Digitalisation can help to make use of this potential and improve sustainability throughout the entire building’s life cycle. One way to address this potential is through the integration of Life-Cycle Assessment (LCA) into the building process by employing Building Information Modelling (BIM). BIM can reduce the effort needed to carry out an LCA and therefore facilitate the integration into the building process. A review of current industry practice and scientific literature shows two main approaches to address BIM-LCA integration. Either the LCA is performed in a simplified way at the beginning of the building process, or it is done at the very end when all the needed information is available, but it is too late for decision-making. One reason for this is the lack of methods, workflows and tools to implement BIM-LCA integration over the entire building process. Therefore, the main objective of this study is to develop an integrated BIM-LCA workflow implemented into a method for the whole building process using an existing structure for cost estimation. A tool is created and used in a case study in Switzerland to test the developed approach. The results of this study show that LCA can be performed continuously in each building phase over the entire building process using existing BIM modelling techniques. The main benefit of this approach is that the re-work caused by the need for re-entering data and the usage of many different software tools that characterise most of the current LCA practices is minimised. Furthermore, decision-making, both at the element and building levels, is supported.

This article presents six probability distributions from the Gamma family with three parameters, for the flood frequency analysis in hydrology. The choice of the Gamma family of statistical dis-tributions was driven by its frequent use in hydrology. In the Faculty of Hydrotechnics, the im-provement of the estimation of maximum flows and including the methodological bases for the realization of a regionalization study with the linear moments method with the corrected pa-rameters was researched, being an element of novelty. The linear moments method is better than MOM because it avoids the choice of skewness depending on the origin of the flows, practiced in Romania. The L-moments method conforms to the current trend for estimating the parameters of statistical distributions. Observed data from hydrometric stations are of relatively short length, so the statistical parameters that characterize them are of a sample that requires correction. The correction of the statistical parameters is proposed, using the method of least squares based on the inverse functions of the statistical distributions expressed with the frequency factor for L-moments. All the necessary elements for their use are presented like, quantile functions, the exact and ap-proximate relations for estimating parameters and frequency factors. A flood frequency analysis case study was carried out for the Ialomita river, to verify the proposed methodology. The per-formance of this distributions is evaluated using Kiling-Gupta and Nash-Sutcliff coefficients.

This study advances the field of solar irradiance nowcasting by introducing a discrete-level classification approach, diverging from traditional continuous measurement methods. Grounded in a need for cost-effective and modular solutions, the research employs high-resolution computer vision coupled with a deep learning framework to predict Direct Normal Irradiance (DNI). By harnessing the capabilities of a Logitech C900 1080p HD camera and an NVIDIA Jetson module, the study achieves real-time data processing, pivotal for CSP systems' operational efficiency. The core of the methodology is the ResNet-50 convolutional neural network, refined via transfer learning on a bespoke dataset, culminating in a predictive accuracy of 85.78%. This discrete classification model contrasts with conventional, costly instruments like the MS-57, offering a novel and accessible alternative for DNI estimation. Such innovation not only demonstrates high predictive accuracy but also signifies a shift towards less resource-intensive and more adaptable solar energy forecasting tools, contributing a significant leap towards optimizing renewable energy systems.