During the last decades, the volume of multimedia content posted in social networks has grown exponentially and such information is immediately propagated and consumed by a significant number of ...users. In this scenario, the disruption of fake news providers and bot accounts for spreading propaganda information as well as sensitive content throughout the network has fostered applied research to automatically measure the reliability of social networks accounts via Artificial Intelligence (AI). In this paper, we present a multilingual approach for addressing the bot identification task in Twitter via Deep learning (DL) approaches to support end-users when checking the credibility of a certain Twitter account. To do so, several experiments were conducted using state-of-the-art Multilingual Language Models to generate an encoding of the text-based features of the user account that are later on concatenated with the rest of the metadata to build a potential input vector on top of a Dense Network denoted as Bot-DenseNet . Consequently, this paper assesses the language constraint from previous studies where the encoding of the user account only considered either the metadata information or the metadata information together with some basic semantic text features. Moreover, the Bot-DenseNet produces a low-dimensional representation of the user account which can be used for any application within the Information Retrieval (IR) framework.
In dynamic and unpredictable environments, the precise localization of first responders and rescuers is crucial for effective incident response. This paper introduces a novel approach leveraging ...three complementary localization modalities: visual-based, Galileo-based, and inertial-based. Each modality contributes uniquely to the final Fusion tool, facilitating seamless indoor and outdoor localization, offering a robust and accurate localization solution without reliance on pre-existing infrastructure, essential for maintaining responder safety and optimizing operational effectiveness. The visual-based localization method utilizes an RGB camera coupled with a modified implementation of the ORB-SLAM2 method, enabling operation with or without prior area scanning. The Galileo-based localization method employs a lightweight prototype equipped with a high-accuracy GNSS receiver board, tailored to meet the specific needs of first responders. The inertial-based localization method utilizes sensor fusion, primarily leveraging smartphone inertial measurement units, to predict and adjust first responders' positions incrementally, compensating for the GPS signal attenuation indoors. A comprehensive validation test involving various environmental conditions was carried out to demonstrate the efficacy of the proposed fused localization tool. Our results show that our proposed solution always provides a location regardless of the conditions (indoors, outdoors, etc.), with an overall mean error of 1.73 m.
Precision cancer medicine suggests that better cancer treatments would be possible guiding therapies by tumor's genomics alterations. This hypothesis boosted exome sequencing studies, collection of ...cancer variants databases and developing of statistical and Machine Learning-driven methods for alterations' analysis. In order to extract relevant information from huge exome sequencing data, accurate methods to distinguish driver and neutral or passengers mutations are vital. Nevertheless, traditional variant classification methods have often low precision in favour of higher recall. Here, we propose several traditional Machine Learning and new Deep Learning techniques to finely classify driver somatic non-synonymous mutations based on a 70-features annotation, derived from medical and statistical tools. We collected and annotated a complete database containing driver and neutral alterations from various public data sources. Our framework, called Driver-Oriented Genomics Analysis (DrOGA), presents the best performances compared to individual and other ensemble methods on our data. Explainable Artificial Intelligence is used to provide visual and clinical explanation of the results, with a particular focus on the most relevant annotations. This analysis and the proposed tool, along with the collected database and the feature engineering pipeline suggested, can help the study of genomics alterations in human cancers allowing precision oncology targeted therapies based on personal data from next-generation sequencing.
Augmented reality can enhance first responders’ situational awareness, displaying data about the environment, location, team status, objectives, and more. However, augmented reality headsets are not ...well suited to operational use, as they are incompatible with personal protective equipment and lack adequate power autonomy. This article presents the smart helmet, a protective helmet featuring an infrared camera, a power source, a processing hub, and a near-eye augmented reality display. The processing hub runs infrared image enhancement, object recognition AI algorithms, and the augmented reality interface, which can be connected to, and display information from, other components. The smart helmet is modular; therefore, individual parts can be selected according to mission needs, including the helmet structure, processing device, additional sensors, and other connected information sources. The whole system is self-reliant and independent from external connectivity. The smart helmet has been tested in three field trials by first responders of diverse sectors.
The continuous evolution of multimedia applications is fostering applied research in order to dynamically enhance the services provided by platforms such as Spotify, Lastfm, or Billboard. Thus, ...innovative methods for retrieving specific information from large volumes of data related with music arises as a potential challenge within the Music Information Retrieval (MIR) framework. Moreover, despite the existence of several musical-based datasets, there is still a lack of information to properly assess an accurate estimation of the impact or the popularity of a song within a platform. Furthermore, the aforementioned platforms measure the popularity in various manners, thus increasing the difficulties in performing generalized and comparable models. In this paper, the creation of SpotGenTrack Popularity Dataset (SPD) is presented as an alternative solution to existing datasets that will facilitate researchers when comparing and promoting their models. In addition, an innovative multimodal end-to-end Deep Learning architecture named as HitMusicNet is presented for predicting popularity in music recordings. Experiments conducted show that the proposed architecture outperforms previous studies in the State-of-the-Art by incorporating three main modalities to the analysis, such as audio, lyrics and meta-data as well as a preliminary compression stage via autoencoder to better the capability of the model when predicting the popularity.
Elderly people care is a major challenge for the smart-cities of future. This represents a valuable opportunity to develop scalable applications to cover the special needs in terms of health ...monitoring and accessibility for people with cognitive impairments. In this paper, a complete system to support daily activities of elderly people based on a multi-sensor scheme is presented. This system is intended to be deployed not only at home, but also at crowded places, such as daily care centers. A multi-layer architecture is drawn to ensure system modularity and interoperability of heterogeneous data with concurrent services. The proposed system includes a set of algorithms for data gathering and processing to detect abnormal events in the considered scenarios. The experiments performed in real scenarios have led to a good performance of the algorithms proposed as well as high accuracy in event detection for both environments.
Indoor Localization and Tracking have become an attractive research topic because of the wide range of potential applications. These applications are highly demanding in terms of estimation accuracy ...and rise a challenge due to the complexity of the scenarios modeled. Approaches for these topics are mainly based on either deterministic or probabilistic methods, such as Kalman or Particles Filter. These techniques are improved by fusing information from different sources, such as wireless or optical sensors. In this paper, a novel MUlti-sensor Fusion using adaptive fingerprint (MUFAF) algorithm is presented and compared with several multi-sensor indoor localization and tracking methods. MUFAF is mainly divided in four phases: first, a target position estimation (TPE) process is performed by every sensor; second, a target tracking process stage; third, a multi-sensor fusion combines the sensor information; and finally, an adaptive fingerprint update (AFU) is applied. For TPE, a complete environment characterization in combination with a Kernel density estimation technique is employed to obtain object position. A Modified Kalman Filter is applied to TPE output in order to smooth target routes and avoid outliers effect. Moreover, two fusion methods are described in this paper: track-to-track fusion and Kalman sensor group fusion. Finally, AFU will endow the algorithm with responsiveness to environment changes by using Kriging interpolation to update the scenario fingerprint. MUFAF is implemented and compared in a test bed showing that it provides a significant improvement in estimation accuracy and long-term adaptivity to condition changes.
Indoor/outdoor localization topic has gained a significant research interest due to the wide range of potential applications. Commonly, the Fingerprinting methods for spatial characterization of the ...environments monitored are employed in deterministic/statistical estimation. However, there are Fingerprint parameters that are generally neglected and can seriously affect the performance yielding to low accurate location. Nowadays, machine and deep learning (DL) methods are employed in this topic due to its ability to approximate complex non-linear models being capable of mitigating the undesirable effects of wireless propagation. In this paper, a complete overview of most influential aspects in Fingerprinting and indoor tracking methods is presented. Furthermore, a novel multi-modal complete tracking system, called SWiBluX, based on statistic and DL techniques is presented. The system relies on relevant feature extraction from available data sources to estimate user's/target indoor position using a multi-phase statistical Fingerprint and DL disruptive approach. In addition, a Gaussian outlier filter is applied to the position estimation model output to further reduce the error in the estimation. The set of experiments performed shows that Fingerprint positioning accuracy estimation can be improved up to 45% resulting in a final estimation error that outperforms related literature.
Indoor person tracking attracts a considerable effort from the research community as it allows to perform Human Behaviour Analysis tasks, where wireless technologies play a key role. However, complex ...signal propagation effects in indoor environments are the main issue to face when performing accurate indoor positioning and tracking. The advances in machine and deep learning models, applied to improve the estimation of the position captured by wireless sensors, can provide a more precise tracking and positioning, an open field for research which has been used to improve the prior art. In this paper, a novel framework for Adaptive Indoor Tracking using Recurrent models, in combination with Generative networks for new data generation (recovery), is presented (RecTrack-GAN). Firstly, a Received Signal Strength Indicator RSSI Fingerprinting database is collected. Secondly, a Recurrent Neural Network (RNN) takes as input the RSSI parameters collected by a Wireless Sensor Network (WSN) and estimates of both orientation and velocity using devices equipped with Inertial Measurement Unit (IMU) sensors, and learns to model the human movement based on these parameters. Thirdly, a Conditional Generative Adversarial Network (CGAN) is used to perform data recovering when no measurements are received and to update the Fingerprinting database taking into account the day time. The experiments performed showed that RecTrack-GAN improves accuracy performance and reduces error deviation for tracking up to 15% compared to the prior art in the literature.
With the technological development in healthcare, environments such as rehabilitation clinics and patients' houses, are increasingly monitored by multi-device systems. To aggregate information, ...overcome privacy issues and device failures, it is essential to match measurements from different sources and associate them to a particular patient. While cameras are used to detect and track anonymized persons, wearable devices can acquire inertial and health information. The challenge is to correctly pair the tracked persons to their status, such as heart rate, collected by other devices. Recently, many works have been proposed, in several scenarios, to tackle sensor fusion-based tracking using a large variety of information. However, when the budget is limited, the involved sensing devices lack of inertial components, such as gyroscope, and may have low precision. In this work, we propose a novel solution to match unlabeled 3D skeletons, detected by a depth camera, with on-wrist wearable devices equipped only with accelerometer. Additionally, a Deep Learning submodule, named SkeletonRNN, is introduced to overcome camera failures in the 3D skeletons points detection and fill missing joints. A complete dataset containing skeletons and accelerations measurements, of dailies and rehabilitation activities, has been collected, manually annotated, and is available for testing purposes. We trained and tested the SkeletonRNN using data augmentation on our dataset, the final average 3D point prediction error is 11.00cm and the skeleton-device pairing accuracy of the overall system is 76.62% on a total of 231 chucks. Datasets, code and experiments can be found at https://github.com/matteo-bastico/SkeletonRNN .
PDF
