Skip to main content

GESTION INTELIGENTE DEL CLOUD CONTINUUM: DESARROLLO DE LAS FUNCIONALIDADES CLAVE DE UN SO

PID2021-124463OB-I00

•
Funding agency name Agencia Estatal de Investigación
Funding agency acronym AEI
Program Programa Estatal para Impulsar la Investigación Científico-Técnica y su Transferencia
Subprogram Subprograma Estatal de Generación de Conocimiento
Call Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Call year 2021
State Plan 2021-2023 Plan Estatal de Investigación Científica y Técnica y de Innovación
Beneficiary centre UNIVERSITAT POLITECNICA DE CATALUNYA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 15
Encontrada(s) 1 página(s)

UPCommons. Portal del coneixement obert de la UPC
Publication Contribución de congreso (ConferenceOutput)

Analyzing distinct neural network models for oxygen saturation prediction towards a personalized COPD management

  • Pascual Saldaña, Heribert|||0000-0003-2114-0929
  • Masip Bruin, Xavier|||0000-0002-4755-556X
  • Alonso Beltran, Albert|||0000-0003-0921-8033
  • Blanco Vich, Isabel
Chronic Respiratory Diseases (CRDs), including Chronic Obstructive Pulmonary Disease (COPD), pose significant global health challenges. Long-term oxygen therapy (LTOT) has been widely adopted to improve respiratory function and quality of life in CRD patients. However, current oxygen dosing approaches show several limitations, including inconvenience and mismatched oxygen flow rates according to patients' activity. To address these challenges, the design of Patient Specific Modeling (PSM) and the utilization of advanced technologies, such as artificial intelligence (AI), can be integrated towards a personalized oxygen dosing. This study presents a first draft of a conceptual architecture combining cloud and edge computing aimed at optimizing oxygen dosing based on individual patient data. The evaluation of sixty neural network architectures in a small-scale test has shown good results, with smaller networks generally outperforming larger ones. Further research is needed to design new testing procedures and explore optimal parameters for the neural networks, including a large patients' number., This work has been supported by the Spanish Ministry of Science and Innovation under grant PID2021-124463OB-I00 funded by MCIN/AEI/10.13039/501100011033 and by ERDF A way of making Europe, the Catalan Government under contract 2021 SGR 00326 and the Catalan Department of Research and Universities, Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Contribución de congreso (ConferenceOutput)

A 3D terrain generator: Enhancing robotics simulations with GANs

  • Arellano García, Silvia
  • Otero Calviño, Beatriz|||0000-0002-9194-559X
  • Kucner, Tomasz Piotr
  • Canal Corretger, Ramon|||0000-0003-4542-204X
Simulation is essential in robotics to evaluate models and techniques in a controlled setting before conducting experiments on tangible agents. However, developing simulation environments can be a challenging and time-consuming task. To address this issue, a proposed solution involves building a functional pipeline that generates 3D realistic terrains using Generative Adversarial Networks (GANs). By using GANs to create terrain, the pipeline can quickly and efficiently generate detailed surfaces, saving researchers time and effort in developing simulation environments for their experiments. The proposed model utilizes a Deep Convolutional Generative Adversarial Network (DCGAN) to generate heightmaps, which are trained on a custom database consisting of real heightmaps. Furthermore, an Enhanced Super-Resolution Generative Adversarial Network (ESRGAN) is used to improve the resolution of the resulting heightmaps, enhancing their visual quality and realism. To generate a texture according to the topography of the heightmap, chroma keying is used with previously selected textures. The heightmap and texture are then rendered and integrated, resulting in a realistic 3D terrain. Together, these techniques enable the model to generate high-quality, realistic 3D terrains for use in robotic simulators, allowing for more accurate and effective evaluations of robotics models and techniques., This work is partially supported by the Spanish Ministry of Science and Innovation under contract PID2021-124463OB-IOO, by the Generalitat de Catalunya under grant 2021-SGR-00326. Finally, the research leading to these results also has received funding from the European Union’s Horizon 2020 research and innovation programme under the HORIZON-EU VITAMIN-V (101093062) project., Peer Reviewed, Objectius de Desenvolupament Sostenible::9 - Indústria, Innovació i Infraestructura
UPCommons. Portal del coneixement obert de la UPC
Publication Contribución de congreso (ConferenceOutput)

A deep learning framework for safety monitoring of a railway section

  • Chriki Zerrouk, Fatima Zohra
  • Simó Mezquita, Ester|||0000-0003-4587-7028
  • Aguiló Gost, Francisco de Asis L.|||0000-0003-1572-8486
  • Garcia Milà, Ignasi
  • Masip Bruin, Xavier|||0000-0002-4755-556X
© 2024 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., This paper presents an approach for anomaly detection and forecasting that aims to protect the integrity and operational continuity of IoT critical infrastructure, specifically for a railway use case. This system is designed to monitor in real-time the sensor data to detect any deviations that may indicate a potential data tampering attack or a sensor malfunction. This early identification and notification of such anomalies is crucial for preventing unauthorized access and mitigating the risks associated with data tampering attacks. The proposed system is comprises of two primary components: a forecasting component and an anomaly detection component. The Forecasting Component uses a Long Short-Term Memory (LSTM) model to predict future sensor values, while the Anomaly Detection Component employs the Tukey's fence method to identify sensor measurements that significantly deviate from normal behaviour. A railway use case is included to demonstrate the practical application of the deep learning framework. These components were evaluated and both demonstrated excellent performance. The Forecasting Component provided highly accurate predictions of future sensor values, while the Anomaly Detection Component effectively identified deviations from normal patterns. The evaluation results confirmed the system's ability to detect significant anomalies and its capability to maintain operational integrity and security in IoT critical infrastructures., The work leading to this publication was partly funded by the European Union’s Horizon Europe under project PHOENi2X, Grant Agreement no. 101070586. For UPC authors this work is partially co-funded by the Spanish Ministry of Science and Innovation under grant PID2021-124463OBI00 funded by MCIN/AEI/10.13039/501100011033 and by ERDF A way of making Europe, as well as by the Catalan Government under contract 2021 SGR 00326., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Contribución de congreso (ConferenceOutput)

Exploring image transformations with diffusion models: A survey of applications and implementation code

  • Arellano García, Silvia
  • Otero Calviño, Beatriz|||0000-0002-9194-559X
  • Tous Liesa, Rubén|||0000-0002-1409-5843
Diffusion Models have become increasingly popular in recent years and their applications span a wide range of fields. This survey focuses on the use of diffusion models in computer vision, specially in the branch of image transformations. The objective of this survey is to provide an overview of state-of-the-art applications of diffusion models in image transformations, including image inpainting, super-resolution, restoration, translation, and editing. This survey presents a selection of notable papers and repositories including practical applications of diffusion models for image transformations. The applications are presented in a practical and concise manner, facilitating the understanding of concepts behind diffusion models and how they function. Additionally, it includes a curated collection of GitHub repositories featuring popular examples of these subjects., This work is partially supported by the Spanish Ministry of Science and Innovation under contract PID2019-107255GB and PID2021-124463OB- IOO, by the Generalitat de Catalunya under grants 2021-SGR-00478 and 2021-SGR-00326. Finally, the research leading to these results also has received funding from the European Union’s Horizon 2020 research and innovation programme under the HORIZON-EU VITAMIN-V (101093062) project., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Contribución de congreso (ConferenceOutput)

Deploying vision foundation AI models on the edge. The SAM2 experience

  • Lin, Zheshuo
  • Tous Liesa, Rubén|||0000-0002-1409-5843
  • Otero Calviño, Beatriz|||0000-0002-9194-559X
As AI-driven applications expand across industries, the need for efficient edge computing solutions becomes increasingly critical. Traditional AI models are designed for high-performance cloud infrastructures, but emerging constraints—such as privacy requirements, network limitations, and real-time processing needs—necessitate optimized deployment on resource-constrained edge devices. This study presents a practical experience in adapting Segment Anything Model 2 (SAM2), a vision foundation model, for edge AI environments. The adaptation process involved translating the model to C++ using ONNX Runtime, enabling efficient execution on heterogeneous hardware. Experimental evaluations demonstrate that deploying SAM2 at the edge enhances processing efficiency, reduces reliance on network stability, and improves real-time responsiveness. This research provides valuable insights into AI in pervasive computing environments, contributing to the sustainable and scalable deployment of foundation models on edge devices., This work is partially funded by the Spanish Ministry of Science and Innovation under contracts PID2019-107255GB and PID2021-124463OB-IOO, as well as by the SGR programs 2021-SGR-00478 and 2021-SGR-00326 of the Catalan government. Finally, the research that led to these results has received funding from the Collaboration Grant for the 2024/2025 period from the Spanish Ministry of Education., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Contribución de congreso (ConferenceOutput)

Malware detection on Linux using runtime opcode tracing

  • Alonso García, Martí
  • Costa Prats, Juan José|||0000-0003-2479-0230
  • Morancho Llena, Enrique|||0000-0003-2403-8145
The fast-paced evolution of cyberattacks to digital infrastructures requires new protection mechanisms to counter-attack them. Malware attacks, a type of cyberattacks ranging from viruses and worms to ransomware and spyware, have been traditionally detected using signature-based methods. But with new versions of malware, this approach is not good enough, and new machine learning tools look promising. In this paper we evaluate five machine learning models (SVM, KNN, Naive Bayes, Decision Tree and Random Forest) to detect Linux malware on applications during its runtime by tracking their executed instructions (opcodes). We show the methodology, the initial dataset preparation, the infrastructure used to obtain the traces of executed instructions, and the evaluation of the results for the different models used. The obtained results show that Random Forest classifier gets the best results, with 90% accuracy or higher., Funded by the European Union. Views and opinions expressed are, however, those of the authors only and do not necessarily reflect those of the European Union or the HaDEA. Neither the European Union nor the granting authority can be held responsible for them. Project number: 101093062 This work is also partially supported by the Spanish Ministry of Science and Innovation under contract PID2021- 124463OB-IOO and by the Generalitat de Catalunya under grant 2021SGR00326., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Contribución de congreso (ConferenceOutput)

High-resolution image generation using artificial intelligence and diffusion modelling

  • Scoles García, Amanda
  • Sionis Ponte, Giulia
  • Otero Calviño, Beatriz|||0000-0002-9194-559X
  • Utrera Iglesias, Gladys Miriam|||0000-0002-0637-1297
Super-Resolution via Repeated Refinement (SR3) is a state-of-the-art super-resolution algorithm based on diffusion model that can enhance the resolution of images. This method is used to pre-trained models on large datasets and can be used for various tasks without requiring training from scratch. Training SR3 from scratch using the ImageNet dataset involves a complex process that requires substantial computational resources and expertise. The idea is applied the trained SR3 model to new images by feeding the low-resolution inputs and obtaining the high-resolution outputs. It's important to note that training SR3 from scratch is a resource-intensive process that requires powerful GPUs and significant computation time. If you do not have access to such resources, an alternative is to use pre-trained models that are already available and fine-tune them on specific datasets or tasks. The paper shows the result of comparing the resolution of the preprocessed images using a significantly smaller number of images to perform the training with those obtained using the pre-trained model. The results obtained show acceptable results without having to perform on large datasets minimizing the computation time to obtain the resolution of images., This work is partially supported by the Spanish Ministry of Science and Innovation under contract PID2021-124463OB-100 and PID2019-107255GB-C22 by the Generalitat de Catalunya under grant 2021-SGR-00326 and 2021-SGR-01007. Finally, the research leading to these results also has received funding from the European Union's Horizon 2020 research and innovation program under the HORIZON-EU VITAMIN-V (101093062) project., Peer Reviewed
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

A deep learning framework for safety monitoring of a railway section

  • Chriki Zerrouk, Fatima Zohra
  • Simó Mezquita, Ester
  • Aguiló Gost, Francisco de Asis L.
  • Garcia Milà, Ignasi
  • Masip Bruin, Xavier
© 2024 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., This paper presents an approach for anomaly detection and forecasting that aims to protect the integrity and operational continuity of IoT critical infrastructure, specifically for a railway use case. This system is designed to monitor in real-time the sensor data to detect any deviations that may indicate a potential data tampering attack or a sensor malfunction. This early identification and notification of such anomalies is crucial for preventing unauthorized access and mitigating the risks associated with data tampering attacks. The proposed system is comprises of two primary components: a forecasting component and an anomaly detection component. The Forecasting Component uses a Long Short-Term Memory (LSTM) model to predict future sensor values, while the Anomaly Detection Component employs the Tukey's fence method to identify sensor measurements that significantly deviate from normal behaviour. A railway use case is included to demonstrate the practical application of the deep learning framework. These components were evaluated and both demonstrated excellent performance. The Forecasting Component provided highly accurate predictions of future sensor values, while the Anomaly Detection Component effectively identified deviations from normal patterns. The evaluation results confirmed the system's ability to detect significant anomalies and its capability to maintain operational integrity and security in IoT critical infrastructures., The work leading to this publication was partly funded by the European Union’s Horizon Europe under project PHOENi2X, Grant Agreement no. 101070586. For UPC authors this work is partially co-funded by the Spanish Ministry of Science and Innovation under grant PID2021-124463OBI00 funded by MCIN/AEI/10.13039/501100011033 and by ERDF A way of making Europe, as well as by the Catalan Government under contract 2021 SGR 00326., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

High-resolution image generation using artificial intelligence and diffusion modelling

  • Scoles García, Amanda
  • Sionis Ponte, Giulia
  • Otero Calviño, Beatriz
  • Utrera Iglesias, Gladys Miriam
Super-Resolution via Repeated Refinement (SR3) is a state-of-the-art super-resolution algorithm based on diffusion model that can enhance the resolution of images. This method is used to pre-trained models on large datasets and can be used for various tasks without requiring training from scratch. Training SR3 from scratch using the ImageNet dataset involves a complex process that requires substantial computational resources and expertise. The idea is applied the trained SR3 model to new images by feeding the low-resolution inputs and obtaining the high-resolution outputs. It's important to note that training SR3 from scratch is a resource-intensive process that requires powerful GPUs and significant computation time. If you do not have access to such resources, an alternative is to use pre-trained models that are already available and fine-tune them on specific datasets or tasks. The paper shows the result of comparing the resolution of the preprocessed images using a significantly smaller number of images to perform the training with those obtained using the pre-trained model. The results obtained show acceptable results without having to perform on large datasets minimizing the computation time to obtain the resolution of images., This work is partially supported by the Spanish Ministry of Science and Innovation under contract PID2021-124463OB-100 and PID2019-107255GB-C22 by the Generalitat de Catalunya under grant 2021-SGR-00326 and 2021-SGR-01007. Finally, the research leading to these results also has received funding from the European Union's Horizon 2020 research and innovation program under the HORIZON-EU VITAMIN-V (101093062) project., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

Exploring image transformations with diffusion models: A survey of applications and implementation code

  • Arellano García, Silvia
  • Otero Calviño, Beatriz
  • Tous Liesa, Rubén
Diffusion Models have become increasingly popular in recent years and their applications span a wide range of fields. This survey focuses on the use of diffusion models in computer vision, specially in the branch of image transformations. The objective of this survey is to provide an overview of state-of-the-art applications of diffusion models in image transformations, including image inpainting, super-resolution, restoration, translation, and editing. This survey presents a selection of notable papers and repositories including practical applications of diffusion models for image transformations. The applications are presented in a practical and concise manner, facilitating the understanding of concepts behind diffusion models and how they function. Additionally, it includes a curated collection of GitHub repositories featuring popular examples of these subjects., This work is partially supported by the Spanish Ministry of Science and Innovation under contract PID2019-107255GB and PID2021-124463OB- IOO, by the Generalitat de Catalunya under grants 2021-SGR-00478 and 2021-SGR-00326. Finally, the research leading to these results also has received funding from the European Union’s Horizon 2020 research and innovation programme under the HORIZON-EU VITAMIN-V (101093062) project., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

An NIDS for known and zero-day anomalies

  • Hussain, Ayaz
  • Aguiló Gost, Francisco de Asis L.
  • Simó Mezquita, Ester
  • Marín Tordera, Eva
  • Masip Bruin, Xavier
Rapid development in the network infrastructure has resulted in sophisticated attacks which are hard to detect using typical network intrusion detection systems (NIDS). There is a strong need for efficient NIDS to detect these known attacks along with ever-emerging zero-day exploits. Existing NIDS are more focused on detecting known attacks using supervised machine learning approaches, achieving better performance for known attacks but poor detection of unknown attacks. Many NIDS have utilized the unsupervised approach, which results in better detection of unknown anomalies. In this paper, we proposed a Hybrid NIDS based on Semisupervised One-Class Support Vector Machine (OC-SVM) and Supervised Random Forest (RF) algorithms. This detection system has several stages. The First stage is based on OC-SVM, which filters benign and malicious traffic. The next stages use many parallel supervised models and an additional OC-SVM model to separate known and unknown attacks from malicious traffic. The previous process is done so that known attacks are classified by their type, and unknown attacks are detected. The proposed NIDS is tested on the standard public dataset CSE-CIC-IDS-2018. The evaluation results show that the system achieves a high accuracy, 99.45%, for detecting known attacks. Our proposed NIDS achieves an accuracy of 93.99% for unknown or zero-day attacks. The overall accuracy of the proposed NIDS is 95.95%. The system significantly improves the detection of known and unknown anomalies using a hybrid approach., This project has received funding from the European Union’s H2020 research and innovation programme under the grant agreement No. 952644, from the Spanish Ministry of Science and Innovation under contract PID2021-124463OB- I00, and from the Catalan Government under contract 2021 SGR 00326., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

Analyzing distinct neural network models for oxygen saturation prediction towards a personalized COPD management

  • Pascual Saldaña, Heribert
  • Masip Bruin, Xavier
  • Alonso Beltran, Albert
  • Blanco Vich, Isabel
Chronic Respiratory Diseases (CRDs), including Chronic Obstructive Pulmonary Disease (COPD), pose significant global health challenges. Long-term oxygen therapy (LTOT) has been widely adopted to improve respiratory function and quality of life in CRD patients. However, current oxygen dosing approaches show several limitations, including inconvenience and mismatched oxygen flow rates according to patients' activity. To address these challenges, the design of Patient Specific Modeling (PSM) and the utilization of advanced technologies, such as artificial intelligence (AI), can be integrated towards a personalized oxygen dosing. This study presents a first draft of a conceptual architecture combining cloud and edge computing aimed at optimizing oxygen dosing based on individual patient data. The evaluation of sixty neural network architectures in a small-scale test has shown good results, with smaller networks generally outperforming larger ones. Further research is needed to design new testing procedures and explore optimal parameters for the neural networks, including a large patients' number., This work has been supported by the Spanish Ministry of Science and Innovation under grant PID2021-124463OB-I00 funded by MCIN/AEI/10.13039/501100011033 and by ERDF A way of making Europe, the Catalan Government under contract 2021 SGR 00326 and the Catalan Department of Research and Universities, Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

A 3D terrain generator: Enhancing robotics simulations with GANs

  • Arellano García, Silvia
  • Otero Calviño, Beatriz
  • Kucner, Tomasz Piotr
  • Canal Corretger, Ramon
Simulation is essential in robotics to evaluate models and techniques in a controlled setting before conducting experiments on tangible agents. However, developing simulation environments can be a challenging and time-consuming task. To address this issue, a proposed solution involves building a functional pipeline that generates 3D realistic terrains using Generative Adversarial Networks (GANs). By using GANs to create terrain, the pipeline can quickly and efficiently generate detailed surfaces, saving researchers time and effort in developing simulation environments for their experiments. The proposed model utilizes a Deep Convolutional Generative Adversarial Network (DCGAN) to generate heightmaps, which are trained on a custom database consisting of real heightmaps. Furthermore, an Enhanced Super-Resolution Generative Adversarial Network (ESRGAN) is used to improve the resolution of the resulting heightmaps, enhancing their visual quality and realism. To generate a texture according to the topography of the heightmap, chroma keying is used with previously selected textures. The heightmap and texture are then rendered and integrated, resulting in a realistic 3D terrain. Together, these techniques enable the model to generate high-quality, realistic 3D terrains for use in robotic simulators, allowing for more accurate and effective evaluations of robotics models and techniques., This work is partially supported by the Spanish Ministry of Science and Innovation under contract PID2021-124463OB-IOO, by the Generalitat de Catalunya under grant 2021-SGR-00326. Finally, the research leading to these results also has received funding from the European Union’s Horizon 2020 research and innovation programme under the HORIZON-EU VITAMIN-V (101093062) project., Peer Reviewed, Objectius de Desenvolupament Sostenible::9 - Indústria, Innovació i Infraestructura, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

Deploying vision foundation AI models on the edge. The SAM2 experience

  • Lin, Zheshuo
  • Tous Liesa, Rubén
  • Otero Calviño, Beatriz
As AI-driven applications expand across industries, the need for efficient edge computing solutions becomes increasingly critical. Traditional AI models are designed for high-performance cloud infrastructures, but emerging constraints—such as privacy requirements, network limitations, and real-time processing needs—necessitate optimized deployment on resource-constrained edge devices. This study presents a practical experience in adapting Segment Anything Model 2 (SAM2), a vision foundation model, for edge AI environments. The adaptation process involved translating the model to C++ using ONNX Runtime, enabling efficient execution on heterogeneous hardware. Experimental evaluations demonstrate that deploying SAM2 at the edge enhances processing efficiency, reduces reliance on network stability, and improves real-time responsiveness. This research provides valuable insights into AI in pervasive computing environments, contributing to the sustainable and scalable deployment of foundation models on edge devices., This work is partially funded by the Spanish Ministry of Science and Innovation under contracts PID2019-107255GB and PID2021-124463OB-IOO, as well as by the SGR programs 2021-SGR-00478 and 2021-SGR-00326 of the Catalan government. Finally, the research that led to these results has received funding from the Collaboration Grant for the 2024/2025 period from the Spanish Ministry of Education., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

Malware detection on Linux using runtime opcode tracing

  • Alonso García, Martí
  • Costa Prats, Juan José
  • Morancho Llena, Enrique
The fast-paced evolution of cyberattacks to digital infrastructures requires new protection mechanisms to counter-attack them. Malware attacks, a type of cyberattacks ranging from viruses and worms to ransomware and spyware, have been traditionally detected using signature-based methods. But with new versions of malware, this approach is not good enough, and new machine learning tools look promising. In this paper we evaluate five machine learning models (SVM, KNN, Naive Bayes, Decision Tree and Random Forest) to detect Linux malware on applications during its runtime by tracking their executed instructions (opcodes). We show the methodology, the initial dataset preparation, the infrastructure used to obtain the traces of executed instructions, and the evaluation of the results for the different models used. The obtained results show that Random Forest classifier gets the best results, with 90% accuracy or higher., Funded by the European Union. Views and opinions expressed are, however, those of the authors only and do not necessarily reflect those of the European Union or the HaDEA. Neither the European Union nor the granting authority can be held responsible for them. Project number: 101093062 This work is also partially supported by the Spanish Ministry of Science and Innovation under contract PID2021- 124463OB-IOO and by the Generalitat de Catalunya under grant 2021SGR00326., Peer Reviewed, Postprint (author's final draft)