DAUIN Ph.D. Day 2024

Immagine
PhdDAUIN24

On Wednesday, November 6, 2024, on the second floor of the Department of Control and Computer Engineering (DAUIN), the Ph.D. students enrolled in the 2nd and 3rd year of the Ph.D. Program in Computer and Control Engineering (37th and 38th cycle) will have a chance to present and discuss in person their research activities in the context of the 8th edition of the DAUIN Ph.D. Day. The event will be an opportunity to illustrate the Ph.D. Programs managed by the Department and disseminate the research activities that are carried out by its staff. The students will be available to share their experience and research activities through posters and demo presentations.

 

 

Agenda

Welcome (9:30 – 9:50)

Prof. Luca Sterpone, Head of the Department of Control and Computer Engineering
Prof. Stefano Grivet Talocia, Head of the Doctoral School 
Prof. Fabrizio Lamberti, Coordinator of the PhD Programme in Computer and Control Engineering
Prof. Stefano di Carlo, Coordinator of the National PhD Programme in Artificial Intelligence – Industry

PhD quality awards: The top PhD students 2024 will be recognized for their outstanding achievements and results.

  • 36th cycle: DAGHERO Francesco, PLIZZARI Chiara, GUERRERO BALAGUERA Juan David
  • 37th cycle: AMPRIMO Gianluca, ANGI Antonino, BERTON Gabriele Moreno, DRI Emanuele, FULCINI Tommaso, MONACO Simone, RANDO MAZZARINO Pietro, RISSO Matteo
Testimonials (9:50 – 10:00)

Dr. Luigi Borzì, former student (35th cycle), now a Research Assistant (RTD-A) at the Department of Control and Computer Engineering
Dr. Eliana Giovannitti, former student (34th cycle), now Analytics Engineer at COMAU

Ph.D. Poster/Demo presentations and discussion (10:00 – 12:45)

During the session, professors and past Ph.D. students will be available to provide information about the Ph.D. Program, admission, as well as formative and research activities.

Awards ceremony (12:45 – 13:00)

Best poster and best demo awards ceremony.

37th Cycle

ALAMIN Khaled Sidahmed Sidahmed

Immagine
ALAMIN

Innovative AI-Enabled Digital Twins for Smart Battery Management: Enhancing Battery Life and Performance in Smart Vehicles

This research advances battery management systems (BMS) by using AI-driven digital twin technology. It combines data-driven methods, machine learning, and cloud-edge connectivity to continuously update models of battery health (SOH) and charge (SOC). This approach optimizes battery performance in real-time, improving electric vehicle reliability and longevity, transforming battery management for software-defined vehicles.

ALI MOUHAMED ALI Romisaa

Immagine
AL

Advanced Navigation Control Systems for Autonomous Vehicles

Our research focuses on developing advanced control systems for mobile robots to navigate from a starting position to a goal position in environments with complexities. Using deep reinforcement learning, we aim to optimize the robot's path to maximize success rates, minimize collisions, and ensure timely navigation. This involves creating algorithms that enable the robot to learn and adapt to dynamic changes in real-time. Our goal is to contribute to safer and more efficient autonomous transportation systems, pushing the boundaries of what autonomous vehicles can achieve in complex environments.

ALMHAITHAWI Doaa

Immagine
ALMHAITHAWI

Exploring The Latent Space for Machine Learning Applications - Security Challenges

The research focuses on exploring latent spaces to improve understanding of AI models' behavior. The main directions are the Theory of building these spaces as estimation of real-world data, the Methodologies to explore their properties (such as projection, exploration, visualization, and others) and the possible real-world Applications to generalize the obtained results.

AMPRIMO Gianluca

Immagine
AMPRIMO

Vision-based, AI-powered approaches to motor functions characterization and rehabilitation

This PhD project explores the intersection of computer vision and artificial intelligence for the automatic and  quantitative characterization of human movement using video-based body pose estimation. Additionally, these technologies are used as human-computer interaction means in exergames for gamified rehabilitation and assessment, suitable for telemedicine systems.

ANGI Antonino

Immagine
ANGI

Integration of ML and NFV for the Management of Distributed Edge Infrastructures

My research focuses on integrating Natural Language Processing, Large Language Models, and Machine Learning algorithms with Software Defined Networks and Intent-based Networks. The goal is to enhance network automation and intelligence using protocols such as OpenFlow and data-plane programming languages such as P4, also allowing more efficient, responsive, and adaptable management of the network.

ANGIONE Francesco

Immagine
ANGIONE

System-Level-Test Techniques for Automotive SoCs

This research investigates System-Level Test (SLT) functional techniques for Automotive SoCs during manufacturing test flow for further increasing quality of semiconductor products. It proposes methodologies for evaluating SLT techniques using a comprehensive test setup that spans various design stages (layout, RTL, code) and data sources (silicon measurements, instruction traces, high-level models). The focus extends beyond core processors to encompass peripherals like timers, communication interfaces, and machine learning accelerators, while also exploring crossbar-specific testing approaches.

BELLARMINO Nicolò

Immagine
BELLARMINO

Computational Intelligence for Computer-Aided Design

Research focuses on developing machine learning (ML)-based methods for microcontroller performance screening. Traditional approaches, based on system-level tests, are time-consuming and limited by application diversity. By using indirect performance monitoring structures like ring oscillators, coupled with ML, it is possible to accurately estimate chip performance from data, ensuring efficient and reliable screening.

BENEDETTO Irene

Immagine
BENEDETTO

Enhancing Decision-Making with Large Language Models in the Legal Domain

Her research focuses on developing innovative solutions in the legal field to enhance content accessibility and expedite decision-making processes using large language models (LLMs). This includes the advancement of state-of-the-art methodologies for named entity recognition, document classification, Explainable AI (XAI), and reasoning applied to the legal domain.

BERTON Gabriele Moreno

Immagine
BERTON

Large-scale visual place recognition

The goal of my research is to be able to confidently and quickly identify the location depicted within a given photo, by comparing it with a database of images of known locations. Such research finds applications in robotics, 3D reconstructions, astronaut photography localization and disaster management.

BUSSA Simone

Immagine
BUSSA

Formal Techniques for Cyber-Physical Systems

The goal of my PhD is to explore the potential of formal methods when applied to cybersecurity management and security verification of cyber-physical systems. On the one hand, I focus on solutions to automatically allocate and configure appropriate security mechanisms to protect a CPS. On the other hand, I focus on the formal verification of these systems from a security and privacy perspective.

DRI Emanuele

Immagine
DRI

Quantum Computing for Finance

The research work explores leveraging quantum computing (QC) to solve complex financial problems. We aim to analyze and improve current quantum finance algorithms to make them practical for real-world use, partnering with Intesa Sanpaolo and IBM Consulting. Key results include enhanced credit risk analysis, practical implementation of quantum PCA, and extending quantum derivative pricing algorithms.

FANÌ Eros

Immagine
FANI

Addressing Heterogeneity in Federated Learning for Real-world Vision Applications

Federated Learning enables training models on edge devices while preserving the user’s privacy. However, non-IID data can lead to biased model training and slow convergence speed. This research addresses the FL challenges in real-world vision applications, identifying the factors that hinder the training process and the potential issues related to the specific tasks and scenarios.

FULCINI Tommaso

Immagine
FULCINI

Gamification of E2E Software

Gamification is defined as the application of game elements in a non-gaming context. Software testing is a discipline often considered boring and uninspiring. This PhD program aims to create a playful environment to support testers during the testing activities, spanning from unit testing to End-to-end testing, to improve their UX by achieving higher user involvement and better performance.

GALANTINO Stefano

Immagine
GALANTINO

Distributed orchestration in a computing continuum

Edge and fog computing are bringing the advantages of cloud technologies at the edge of the network. In the near feature we expect to include also customer devices in this computing continuum, creating the so called “liquid computing” in which tasks can be executed almost anywhere. The current PhD is oriented to investigate the continuum and its applications with a focus on sustainable computing and energy and battery management.

GIANNANTONI Leonardo

Immagine
GIANNANTONI

From Biological Inspiration to AI-Driven Algorithms and Back: Modeling, Simulation, and Optimization of Complex Systems

Leonardo specializes in AI-driven modeling, simulation, and optimization of complex systems. His expertise includes co-simulation frameworks, reinforcement learning, formal languages, and computational biology, with applications in synthetic biology, biofabrication, and structural biology. His recent research focuses on advancing AI-based explainability in protein folding.

GOLETTO Gabriele

Immagine
GOLETTO

Development of sustainable computer vision algorithms on edge computing devices

Egocentric vision, a specialized computer vision field, analyzes images and videos from wearable cameras. Despite its promising applications in wearable devices, it faces challenges like online processing, environmental bias, and computational cost in real-world settings. My research aims to address these issues to advance egocentric vision for practical use.

IARIA Giusy

Immagine
IARIA

Towards Ultra-Reliable Automotive SoC

Safety-critical devices require testing techniques to ensure their reliability during the entire lifetime. The research aims to integrate and innovate such techniques for modern complex Automotive SoCs, with two goals: decreasing the testing time and costs without losing reliability and collecting on-chip information during the life of the embedded systems to diagnose faulty behaviors of field returns.

INSINGA Giorgio

Immagine
INSINGA

Test and diagnosis of memories embedded in Automotive SoC

My research focuses on the reliability of Automotive SoCs. In particular, I have created a series of algorithms for the compaction and compression of diagnostic data generated during the SoC memory test. I have also developed a method to reconstruct the memory characteristics starting from the irradiation test data and a new type of test-oriented ADC converter.

MARTINI Lorenzo

Immagine
MARTINI

Complex systems science and brain dynamics

Many people may remember the phrase “DNA is the blueprint of life.” But if all human cells share the same blueprint, where does the incredible cell diversity come from? The answer lies in specific gene expression profiles and the complex system of transcriptional regulation. My work focuses on creating computational tools and bioinformatic workflows to investigate and integrate cutting-edge experimental data to shed light on brain dynamics process and cancer research.

MONACO Simone

Immagine
MONACO

Theory-inspired deep learning

My research combines the benefits of deep learning, which is generalizable but data-intensive, with theory-based models relying on prior knowledge but limited in some real-world scenarios. This involves using various training strategies and enhancing traditional architectures. The goal is to improve efficiency and model convergence speed, creating more expressive tools to explain model outcomes.

PEGORARO CHENET Cristiano

Immagine
PEGORARO

Hardware Based Neuromorphic Resilience to Faults and Attacks

Code anomalies may affect modern computing systems’ safety, reliability, and security. They include bugs, malicious code, and indirect effects caused by faults in the physical layer. This Ph.D. project aims to investigate the capabilities of hardware events and artificial intelligence to improve resilience to code anomalies. This approach has advantages like real-time detection and minimal performance overhead.

PIANO Luca

Immagine
PIANO

Advanced learning strategies for visual understanding

Developing more fair and robust deep neural networks is my doctoral challenge. By integrating prior knowledge, I guide the learning of DNNs to overcome several key obstacles: limited labeled data, bias, shortcuts, and increase their controllability. My projects include a latent diffusion model-based system for anonymizing images, preserving the original content but removing identifying information.

PIRRERA Simone

Immagine
PIRRERA

Optimization methods for dynamical system learning

Recurrent neural networks (RNNs) are a popular tool for learning dynamical systems from data. However, training RNNs is widely recognized as challenging due to vanishing gradient issues. To train RNNs effectively, we investigate novel problem formulations and optimization algorithms. To achieve this, we propose a novel paradigm for constrained optimization based on feedback control theory.

PORTALURI Andrea

Immagine
PORTALURI

AI-oriented Design Methods for Intelligent Aerospace Computing Systems

Automatization and AI in satellite computing are revolutionizing modern technology, allowing satellites to make autonomous decisions without ground control, greatly benefiting space missions. However, AI's use in space is limited due to unknowns about its behavior in harsh environments. My research focuses on AI algorithm pipelines in space, aiming to develop CAD methods for designing resilient AI algorithms and effectively mapping them onto modern edge reconfigurable devices.

RANDO MAZZARINO Pietro

Immagine
RANDO

Simulation and Modelling of Energy Demand Flexibility in energy communities

My research aims to explore solutions in the energy sector to enhance flexibility and operational efficiency. Specifically, I began by automating methods for modeling multi-energy systems, followed by creating a standardized co-simulation framework for their simulation. Finally, I tested various machine learning-based strategies to optimize their control. These efforts contribute to advancing the integration and management of complex energy systems.

RISMA CARLETTI Carlos Mateo

Immagine
RISMA

Vehicular Micro-Cloud Architectures and Applications

Future Connected Autonomous Vehicles (CAVs) generate and share vast amounts of data. Efficient sharing and processing of redundant data created by neighboring CAVs is crucial. Vehicular Micro Clouds (VMCs), where clusters of vehicles (such as platoons) collaborate via vehicular networks, offer a promising solution. This research seeks new methods to efficiently aggregate context data within VMCs.

RISSO Matteo

Immagine
RISSO

Automatic Hardware-aware Design and Optimization of Deep Learning Models

Deep Neural Networks (DNNs) popularity requires optimizing them for many use-cases, especially for edge devices with limited resources. Manual tuning is suboptimal, leading to Neural Architecture Search (NAS). However, NAS has limitations in complexity and hardware awareness. This thesis addresses these issues with PLiNIO, an open-source framework for efficient, hardware-aware DNN optimization.

ROBBIANO Luca

Immagine
ROBBIANO

Hardware-oriented Neural Architecture Search

In the last few year, Neural Architecture Search (NAS) algorithms have emerged to automate neural network design. This project aims to solve this issue by developing scalable NAS methods able to satisfy hardware-related requirements and ease the development and deployment of efficient deep learning models for real-world applications, from embedded devices used in robotics to large-scale clusters.

SAEEDI Sepide

Immagine
SAEEDI

Fast and Accurate Prediction of the Impact of Approximate Operators on a Complex Computation

Approximate Computing (AxC) techniques trade off accuracy for performance and energy efficiency. To address the problem of finding optimal AxC techniques for each application, we proposed 2 Design Space Exploration approaches: 1- an Interval Arithmetic-based approach to rapidly evaluate the approximation error. 2- a Reinforcement Learning approach to efficiently explore a vast design space.

SISINNI Silvia

Immagine
SISINNI

Zero-Trust Security of Network Nodes

My research focuses on defining Zero-Trust architectures to secure modern ICT infrastructures, which now involve distributed data processing across edge, fog, and cloud computing. I explore techniques for device identity verification, secure node configuration, and verification of the security posture throughout the entire lifecycle of each component, ensuring robust security in dynamic environments.

TRIVIGNO Gabriele

Immagine
TRIVIGNO

Attention guided cross domain geolocalization

Geolocation is a challenging task as photos may offer few cues about their location. Traditional algorithms rely on the trained features The problem is challenging when there is a shift between training and test data distribution. My research studies the problem of geo-localization across visual domains. We combine attention mechanisms with modern domain adaptation algorithms.

VACCA Eleonora

Immagine
VACCA

Multi-Test and Validation Platform for AI-oriented Sensor System Integration

AI reliability is crucial for safety-critical application. My research focuses on developing reliable AI accelerators. Specifically, I designed zero-overhead self-test methods and advanced algorithm-based fault mitigation. My work has been applied in industry, including the Zodiac Pioneer Space Mission and the TERRAC project with the European Space Agency.

VAIANI Lorenzo

Immagine
VAIANI

Exploring the use of deep models in multimodal scenarios

My PhD research focuses on advancing multimodal learning, particularly in vision-language tasks. I analyze social media content to detect hate speech and fake news, and work on Visually Rich Document Understanding through VQA and key information extraction. Additionally, I explore multimodal tasks involving audio, such as podcast summarization and video emotion recognition.

38th Cycle

AHMED ELTARAS Tamer

Immagine
AHMED

Enhancing Security and Privacy in Federated Learning through Advanced Adversarial Analysis

Federated learning has emerged as a promising approach for preserving privacy. However, recent studies show that private training data can be leaked through many attacks. Our research examines these security weaknesses, focusing on how adversaries extract private information from gradients, with the goal of developing comprehensive solutions to protect data privacy in federated learning.

AVIGNONE Andrea

Immagine
AVIGNONE

Exploring AI solutions to unveil Urban Data

The future of smart cities relies on the combination of urban data science and Artificial Intelligence, which are crucial for enhancing citizens' quality of life and integrating new technologies. This research exploits advanced machine learning and deep learning techniques to transform complex and heterogeneous data into actionable insights, providing effective solutions to real-world challenges.

BAROCCI Michelangelo

Immagine
BAROCCI

Design of digital/neuromorphic computing systems and SDK

This PhD research, funded by EBRAINS-Italy, aims to accelerate the engineering process behind the simulation of Spiking Neural Networks through the creation of heterogeneous SoCs on FPGA-based systems. The system will support the development of neuromorphic applications by combining the advanced computing capabilities of SNNs accelerators and the configurability provided by traditional processors.

BERGAMASCO Letizia

Immagine
BERGAMASCO

Early markers to predict the evolution from Mild Cognitive Impairment to Overt Dementia, and support to differential diagnosis of diverse forms of dementia

Dementia is a decline in cognitive function and the leading cause of disability in the elderly. Early diagnosis is challenging but crucial to improve quality of life. This work investigates facial expressions as possible early markers of dementia. Artificial intelligence methods for facial expression analysis are explored to support early and differential diagnosis of different types of dementia.

BRAVI Enrico

Immagine
BRAVI

Security of cyber-physical systems

My research focuses on securing devices by designing and implementing solutions leveraging principles like Trusted Computing. This involves evaluating hardware and software support for trusted execution on limited-capability nodes. The goal is to demonstrate a secure CPS ecosystem with strong identity, secure communications, and integrity monitoring using Trusted Computing and AI techniques.

BUTERA Alberto

Immagine
BUTERA

Blockchain Technology for Intelligent Vehicles

As a second-year PhD student, I am advancing blockchain technology for smart vehicles. My research develops innovative distributed ledger solutions to enhance vehicle data traceability, autonomous system integration, and cross-platform interoperability. By leveraging cutting-edge blockchain protocols, I aim to revolutionize smart transportation addressing key challenges for intelligent vehicles.

CARDONE Lorenzo

Immagine
CARDONE

GPU-based Parallel Techniques for the Evaluation of Reliability Measures of Large System-on-Chip devices

Graphs can be used to represent almost any problem and are therefore used in a multitude of fields. This research focuses on understanding the intricacies of hard graph problems and the main parallelization schemes applied to them. These techniques will be applied to the implementation of a parallel logic simulator that is able to exploit the power of many-core architectures.

CUTTANO Claudia

Immagine
CUTTANO

Enabling complex computer vision tasks on the edge: applications to semantic segmentation and beyond

Understanding scenes is crucial for intelligent systems to interpret and interact with their environments. This encompasses tasks like image segmentation (classifying pixels into categories) and referring segmentation (identifying objects from textual descriptions). My research focuses on developing robust models to handle real-world scene complexity, emphasizing efficient, real-time solutions.

D'AGOSTINO Pietro

Immagine
D'AGOSTINO

Innovative solutions for Industrial IoT

The goal of the research is to develop solutions to apply the concept of Industrial Internet of Things (IIoT) to machinery in the beverage and food industry. I analyze industrial needs to define the challenges and problems to be solved. I collaborate with AROL to understand the requirements of the bottling sector and define a reference architecture for an IIoT solution that includes field, edge, cloud, and connectivity, with a focus on functionalities such as storage and processing.

DE SANTIS Francesco

Immagine
DE

Explainable Large Language Models

Large Language Models (LLMs) excel in text understanding and generation. However, their complexity makes it difficult to understand the decision path they took to make a prediction or to generate text. Therefore, it is crucial to develop tools that can reveal model behaviour, ensuring more reliable LLMs.

DILILLO Nicola

Immagine
DILILLO

Analysis of Multispectral Plant Imaging Data

Multispectral imagery can revolutionize disease prediction methods and plant phenotyping in agriculture. My research aims to identify the most effective wavelengths, improving prediction accuracy and simplifying the  classification process by reducing the number of variables used.  This approach will lead to developing more cost-effective and efficient tools for these applications.

FILIPPONI Gabriele

Immagine
FILIPPONI

Manufacturing and in-field testing techniques

This research aims to enhance the reliability of modern SoC in automotive and safety-critical sectors. Current testing methods fall short in ensuring reliability, necessitating new approaches covering a device's lifecycle. Key areas include improving functional manufacturing testing to reduce yield loss, developing in-field self-test routines for high fault coverage, and using logic diagnosis.

FIORENZA Jacopo

Immagine
FIORENZA

XR for Cultural Heritage Accessibility

The topic of eXtended Reality (XR) for Cultural Heritage (CH) is highly relevant in the scientific community, particularly in computer science. The number of works combining CH and XR has surged, covering virtual exhibitions, reconstructions, narrative experiences, and educational systems. Accessibility in XR and CH is either implicitly or explicitly addressed, but deeper exploration is needed. Key elements are physical and perceptual accessibility, catering to users with disabilities or neurological disorders.

FONTANA CRESPO Rafael Natalio

Immagine
FONTANA

Digital Twin Models for Marine Offshore Renewable

The PhD activities focus on the design and development of Digital Twin of real-world Wave Energy Converters (WECs) to unlock algorithms both i) to forecast energy production in the short and mid-term and ii) to promote predictive maintenance strategies. The objectives also comprise the development of an Internet of Things (IoT) platform able to collect, store and retrieve the WEC’s sensed data.

FOSCALE Tommaso

Immagine
FOSCALE

Testing Techniques for Automotive Systems on Chip

The proposed research project will focus on testing techniques used for automotive systems-on-chip. The study aims to identify the most effective testing methods to recognize chip defects and reduce the number of defective microelectronic devices entering the market. The outcomes of this study will contribute to the creation of new testing techniques that can be applied to microelectronic devices.

GALLIPOLI Giuseppe

Immagine
GALLIPOLI

Conversational AI – New tasks and models for conversational applications

Given the rapid proliferation of Conversational AI across academia and industry and its widespread user adoption, my research aims to explore the diverse tasks and models within this field. By leveraging the unique characteristics of conversational data, my goal is to develop more effective solutions for both new applications and traditional NLP tasks applied to the conversational domain.

GARACCIONE Giacomo

Immagine
GARACCIONE

Advancing UI testing techniques

My PhD research focuses on adopting gamification (using game-like elements in non-recreational concepts for increased enjoyment) to improve two software engineering topics: requirements definition through software modeling, and user interface testing. For the former, I have defined a framework for gamifying modeling education, and for the latter I have developed IDE and web plugins for UI testing.

GATTI Gabriele

Immagine
GATTI

Automatic Analysis of Cybersecurity Risks

The research aims at developing methodologies for the collection and evaluation of cyber risk related information so that companies with low technical skills and resources can understand their exposure to cyber threats. The research started with the development of an automated system that can collect publicly available information, assign it a score and aggregate these evaluations into summarizing indexes for ease of comprehension. The current focus is on the validation of the scoring system and the exploration of how AI can support this process.

KOLAHIMAHMOUDI Nima

Immagine
KOLAHIMAHMOUDI

Functional Safety Techniques for Automotive-oriented Systems-on-Chip

With recent improvements in System-on-Chip (SoC) technologies, the complexity of SoCs is advancing, and their dimensions are shrinking, so they have more frequent physical defects. Automotive SoCs are used in fields such as electric vehicles, aerospace, etc.; testing is needed for high application quality. The main target of my research is to develop self-testing circuits inside SoCs.

KOUDOUNAS Alkis

Immagine
KOUDOUNAS

Explainable AI for spoken language understanding

End-to-end spoken language understanding (SLU) models infer semantic information from spoken signals, improving accuracy over traditional SLU models. Their black-box nature lacks in explainability and transparency. Collaborating with Amazon AGI, my research focuses on explaining SLU model prediction, identifying subgroups with anomalous behavior, and enhancing performance while mitigating biases.

LIMAS SIERRA Robert Alexander

Immagine
LIMAS

New Techniques for the Reliability Evaluation of Neural Networks running on GPUs

The reliability of neural networks deployed on GPUs is crucial in many domains (automotive). Given the complexity of both hardware and software, estimating reliability figures is challenging. My activities focus on new approaches to characterize the vulnerability to faults affecting the hardware while exploring methods to detect faults and developing clever strategies to enhance system resilience.

MALAN Erich

Immagine
MALAN

Distributed and Federated Learning over IoT networks

Distributed and Federated Learning enable the sharing of data-driven knowledge among smart sensors under privacy and infrastructure restrictions. Yet, communication bottlenecks and limited device resources challenge the development of scalable and robust distributed learning. Novel and multifaceted solutions are required to efficiently address infrastructure costs, data annotation, and security concerns to unlock the potential of edge intelligence.

MARINO Jacopo

Immagine
MARINO

Efficient 6G Softwarized Network Infrastructures

My research enhances Kubernetes clusters for seamless resource sharing in the Computing Continuum, offloading tasks from robots to 6G edge or cloud resources. This reduces battery use and boosts computational power by accessing advanced resources like GPUs, and idle charging robots to improve efficiency, aiming to create a more efficient computing environment for autonomous robots.

MILAZZO Rosario

Immagine
MILAZZO

Robust, accurate and efficient deep neural networks for satellite and mobility applications

Computing platforms in vehicles and satellites are becoming more powerful. However, deploying deep learning (DL) applications on these platforms requires deep neural networks (DNNs) that maintain high accuracy, efficiency, and robustness, even in the presence of transient faults like radiation-induced errors. This research focuses on ensuring DNNs meet these stringent standards.

NAPOLI Emanuele Antonio

Immagine
NAPOLI

Analysis, search, development, and reuse of smart contracts

My research in blockchain technology focuses on innovative approaches to simplify smart contracts creation. By leveraging Large Language Models and advanced visualization techniques, the objective is to make the complex process of writing smart contracts more accessible and efficient. This innovative approach combines cutting-edge AI with visual representations to enhance blockchain development.

NAPOLITANO Davide

Immagine
NAPOLITANO

Neural Explainers and Multimodal Learning

My PhD research focuses on developing and exploring new neural approaches to enhance the interpretability of complex AI models and advancing multimodal learning techniques to effectively integrate data from diverse sources, such as text, images, and tabular data. By achieving this, I aim to design AI systems able to make existing AI models more transparent and explainable, providing deeper insights into their decision-making processes and fostering greater trust in their applications across various domains

PIGNARI Riccardo

Immagine
PIGNARI

New computational paradigms for neuromorphic HW

The aim of the PhD research activities is to investigate the novel computational paradigm enabled by neuromorphic models, implemented with spiking neural networks, in the areas of machine learning and edge computing. The focus is on digital-to-spike data encoding methods and the adoption of these networks to solve problems belonging to the class of constraint satisfaction problems.

REGE CAMBRIN Daniele

Immagine
REGE

Efficient and Effective Deep Learning for Crisis Management and Ecology

The recent availability of public satellite imagery and textual information are excellent sources to support ecological transition and crisis management. The research focuses on developing efficient and scalable machine-learning solutions to help humanity face great challenges, such as managing natural disasters, preserving green areas, and improving food production.

RONDINA Marco

Immagine
RONDINA

Responsible development and use of AI: from principles to practices

I am on a journey to promote the responsible development and use of AI. My research focuses on creating a metrics-based framework for empirical evaluation of training data. By looking and data quality, balance and documentation, I aim to address ethical challenges in AI. I'm also auditing the outcomes of AI systems to investigate the most common patterns of discrimination.

RUSSO Alessandro Sebastian

Immagine
RUSSO

Neural-symbolic scene integration & generation

Neuro-Symbolic (NeSy) techniques enhance scene understanding by combining neural learning with symbolic logic. As a second-year PhD student, my research focuses on using NeSy in Scene Graph Generation (SGG) to detect objects and refine relationship predictions in visual scenes. Applications include autonomous driving and improved human-computer interaction, combining robustness and clarity.

RUSSO Marco

Immagine
RUSSO M

Quantum Machine Learning

QML explores the applications of quantum computing to machine learning and vice versa. It involves both using quantum algorithms to enhance traditional ML tasks, potentially offering exponential speedups, and applying ML techniques to quantum control, improving the efficiency and performance of quantum operations at the physical level.

VERCELLINO Chiara

Immagine
VERCELLINO

Quantum Machine Learning Applications and Algorithms for Industrial Use Cases

Quantum computing faces challenges such as low qubit counts and limited coherence times, hindering the use of disruptive algorithms. To align the technology with practical industrial applications, I conduct research in hybrid quantum-classical algorithms, specifically focusing on quantum optimization and quantum machine learning on NISQ (Noisy Intermediate-Scale Quantum) devices.

VEZZANI Leonardo

Immagine
VEZZANI

In Car AR HUD

My phd focuses on Augmented Reality on Head-up Displays for automotive in collaboration with Reply, funded by a PNRR scholarship. Using the CARLA driving simulator, I investigate the role of AR HUDs in realistic scenarios. Using such a tool result in a flexible, shareable, and cost-effective approach, creating shareable and innovative solutions for a safer and more intuitive driving experience.

VISCONTI Alessandro

Immagine
VISCONTI

Digital Twin, Presence & Interaction in Metaverses

As a second-year PhD student, my research aims to improve the metaverse, immersive, social and decentralised 3D world. My goal is to enhance user interaction, immersion and presence through realistic representations, intuitive interaction mechanisms, creating authentic virtual experiences and social interactions through cutting-edge technologies and digital twinning solutions.

VITALI Giacomo

Immagine
VITALI

Quantum Machine Learning Applications and Algorithms for Technology Transfer

Quantum Machine Learning holds the potential to outperform classical methods. However, the embryonic stage of these technologies limits the applicability of such algorithms to real-world problems. My work focuses on reducing this gap by matching applications to the most suitable combination of state-of-the-art QML algorithms and technological platform, with a focus on neutral atom simulators.

ZACCONE Riccardo

Immagine
ZACCONE

Algorithms and architectures for distributed training of deep neural networks

Modern success of deep learning relies on increasingly complex models, whose training requires large amounts of data and computing resources. This PhD project investigates the problem of distributing the training of deep neural networks and developing novel algorithms and model architectures to enable efficient and effective decentralized training for next generation deep learning applications.