La responsabilità per attività (d'impresa) pericolosa: nuovi rischi di danno e scelte di politica legislativa nell'ambito dell'U.E.

La tesi analizza il criterio di responsabilità civile fondato sull'esercizio di attività pericolose (art. 2050 c.c.), proponendone un inquadramento all'interno del campo della responsabilità oggettiva. Viene dunque approfondito, in quest'ottica, il problema del contenuto della prova liberatoria. Anche alla luce del confronto con altri ordinamenti giuridici, poi, l'elaborato si sofferma sulla indefinita nozione di pericolosità e sulle possibili strade per circoscriverla. Nell'ultima parte della tesi, ci si intrattiene sulle interferenze tra responsabilità civile disciplinata a livello comunitario (in particolare quella per danni cagionati da prodotti difettosi) e criterio di imputazione fondato sul pericolo. Tale ampia disamina consente infine di affrontare il problema della regolazione dei rischi di danno implicati dallo sviluppo delle tecniche di Intelligenza artificiale. Lo scopo è quello di contribuire al dibattito che è in corso presso la dottrina e presso le Istituzioni dell'Unione europea, dal punto di osservazione del modello nazionale delle "attività pericolose".

Adaptive Neural Translation for Enterprises (ANTE). La traduzione automatica neurale a sostegno delle imprese emiliano-romagnole

Il progetto ANTE riguarda i nuovi sistemi di traduzione automatica (TA) e la loro applicazione nel mondo delle imprese. Lo studio prende spunto dai recenti sviluppi legati all’intelligenza artificiale e ai Big Data che negli ultimi anni hanno permesso alla TA di raggiungere livelli qualitativi molto elevati, al punto tale da essere impiegata da grandi multinazionali per raggiungere nuove quote di mercato. La TA può rispondere positivamente anche ai bisogni delle imprese di piccole dimensioni e a basso tenore tecnologico, migliorando la qualità delle comunicazioni multilingue attraverso delle traduzioni in tempi brevi e a costi contenuti. Lo studio si propone quindi di contribuire al rafforzamento della competitività internazionale delle piccole e medie imprese (PMI) emiliano- romagnole, migliorando la loro capacità di comunicazione in una o più lingue straniere attraverso l’introduzione e l’utilizzo efficace e consapevole di soluzioni ICT di ultima generazione e fornire, così, nuove opportunità di internazionalizzazione.

AI and legal personhood: a theoretical survey

I set out the pros and cons of conferring legal personhood on artificial intelligence systems (AIs), mainly under civil law. I provide functionalist arguments to justify this policy choice and identify the content that such a legal status might have. Although personhood entails holding one or more legal positions, I will focus on the distribution of liabilities arising from unpredictably illegal and harmful conduct. Conferring personhood on AIs might efficiently allocate risks and social costs, ensuring protection for victims, incentives for production, and technological innovation. I also consider other legal positions, e.g., the capacity to act, the ability to hold property, make contracts, and sue (and be sued). However, I contend that even assuming that conferring personhood on AIs finds widespread consensus, its implementation requires solving a coordination problem, determined by three asymmetries: technological, intra-legal systems, and inter-legal systems. I address the coordination problem through conceptual analysis and metaphysical explanation. I first frame legal personhood as a node of inferential links between factual preconditions and legal effects. Yet, this inferentialist reading does not account for the ‘background reasons’, i.e., it does not explain why we group divergent situations under legal personality and how extra-legal information is integrated into it. One way to account for this background is to adopt a neo-institutional perspective and update its ontology of legal concepts with further layers: the meta-institutional and the intermediate. Under this reading, the semantic referent of legal concepts is institutional reality. So, I use notions of analytical metaphysics, such as grounding and anchoring, to explain the origins and constituent elements of legal personality as an institutional kind. Finally, I show that the integration of conceptual and metaphysical analysis can provide the toolkit for finding an equilibrium around the legal-policy choices that are involved in including (or not including) AIs among legal persons.

Continual learning for computer vision applications

One of the most visionary goals of Artificial Intelligence is to create a system able to mimic and eventually surpass the intelligence observed in biological systems including, ambitiously, the one observed in humans. The main distinctive strength of humans is their ability to build a deep understanding of the world by learning continuously and drawing from their experiences. This ability, which is found in various degrees in all intelligent biological beings, allows them to adapt and properly react to changes by incrementally expanding and refining their knowledge. Arguably, achieving this ability is one of the main goals of Artificial Intelligence and a cornerstone towards the creation of intelligent artificial agents. Modern Deep Learning approaches allowed researchers and industries to achieve great advancements towards the resolution of many long-standing problems in areas like Computer Vision and Natural Language Processing. However, while this current age of renewed interest in AI allowed for the creation of extremely useful applications, a concerningly limited effort is being directed towards the design of systems able to learn continuously. The biggest problem that hinders an AI system from learning incrementally is the catastrophic forgetting phenomenon. This phenomenon, which was discovered in the 90s, naturally occurs in Deep Learning architectures where classic learning paradigms are applied when learning incrementally from a stream of experiences. This dissertation revolves around the Continual Learning field, a sub-field of Machine Learning research that has recently made a comeback following the renewed interest in Deep Learning approaches. This work will focus on a comprehensive view of continual learning by considering algorithmic, benchmarking, and applicative aspects of this field. This dissertation will also touch on community aspects such as the design and creation of research tools aimed at supporting Continual Learning research, and the theoretical and practical aspects concerning public competitions in this field.

Synthesis and catalytic activity of new cu-based catalytic systems for the enhanced electrochemical CO2 reduction

This Ph.D. thesis concerns the synthesis of nanostructured Cu-containing materials to be used as electrode modifiers for the CO2 electroreduction in aqueous phase and the evaluation of their catalytic performances. Inspired by the fascinating concept of the artificial photosynthesis-oriented systems, several catalytic layers were electrochemically loaded on carbonaceous gas diffusion membranes, i.e., 3D structures that allow the design of eco-friendly materials for applications in green carbon recycling processes. In particular, early studies on Cu(I-II)-Cu(0) nanostructured materials were carried out to produce films on 4 cm2 sized supports by means of a fast and low-cost electrochemical procedure. Besides, through a screening of potentials, it was possible to find out a selective value for the CH3COOH production at -0.4 V vs RHE with a maximum productivity (1h reaction), ensured by the presence of the Cu+/Cu0 active redox couple (0.31 mmol gcat-1 h-1). On the basis of these results, further optimisations of the electrocatalyst chemical composition were carried out with the aim of (i) facilitating the interaction with CO2, (ii) increasing the dispersion of the catalytic active phase, and (iii) enhancing the CH3COOH productivity. To this aim, novel electrocatalysts based on layered double hydroxides (LDHs) were optimised, having as a final goal the formation of a new Cu2O-Cu0 based electrocatalyst derived from electrochemically achieved CuMgAl LDHs, subjected to calcination and reduction processes. The as-obtained electrocatalysts were tested for the selective production of CH3COOH and unprecedented results were obtained with the pristine CuMgAl LDH (2.0 mmol gcat-1 h-1). Additional characterisations of such an electrocatalyst have highlighted the possibility to achieve a ternary LDH in intimate contact with Cu2O-Cu0 species starting from the electrochemical deposition. The presence of these species, along with an alkaline environment on the electrode surface, were essential to preserve the selectivity towards the desired product, as confirmed by further operando studies.

Applied deep learning and data science with a human-centric and data-centric approach

The term Artificial intelligence acquired a lot of baggage since its introduction and in its current incarnation is synonymous with Deep Learning. The sudden availability of data and computing resources has opened the gates to myriads of applications. Not all are created equal though, and problems might arise especially for fields not closely related to the tasks that pertain tech companies that spearheaded DL. The perspective of practitioners seems to be changing, however. Human-Centric AI emerged in the last few years as a new way of thinking DL and AI applications from the ground up, with a special attention at their relationship with humans. The goal is designing a system that can gracefully integrate in already established workflows, as in many real-world scenarios AI may not be good enough to completely replace its humans. Often this replacement may even be unneeded or undesirable. Another important perspective comes from, Andrew Ng, a DL pioneer, who recently started shifting the focus of development from “better models” towards better, and smaller, data. He defined his approach Data-Centric AI. Without downplaying the importance of pushing the state of the art in DL, we must recognize that if the goal is creating a tool for humans to use, more raw performance may not align with more utility for the final user. A Human-Centric approach is compatible with a Data-Centric one, and we find that the two overlap nicely when human expertise is used as the driving force behind data quality. This thesis documents a series of case-studies where these approaches were employed, to different extents, to guide the design and implementation of intelligent systems. We found human expertise proved crucial in improving datasets and models. The last chapter includes a slight deviation, with studies on the pandemic, still preserving the human and data centric perspective.

Concepts and methods for efficient decentralized learning in federated settings

Deep Neural Networks (DNNs) have revolutionized a wide range of applications beyond traditional machine learning and artificial intelligence fields, e.g., computer vision, healthcare, natural language processing and others. At the same time, edge devices have become central in our society, generating an unprecedented amount of data which could be used to train data-hungry models such as DNNs. However, the potentially sensitive or confidential nature of gathered data poses privacy concerns when storing and processing them in centralized locations. To this purpose, decentralized learning decouples model training from the need of directly accessing raw data, by alternating on-device training and periodic communications. The ability of distilling knowledge from decentralized data, however, comes at the cost of facing more challenging learning settings, such as coping with heterogeneous hardware and network connectivity, statistical diversity of data, and ensuring verifiable privacy guarantees. This Thesis proposes an extensive overview of decentralized learning literature, including a novel taxonomy and a detailed description of the most relevant system-level contributions in the related literature for privacy, communication efficiency, data and system heterogeneity, and poisoning defense. Next, this Thesis presents the design of an original solution to tackle communication efficiency and system heterogeneity, and empirically evaluates it on federated settings. For communication efficiency, an original method, specifically designed for Convolutional Neural Networks, is also described and evaluated against the state-of-the-art. Furthermore, this Thesis provides an in-depth review of recently proposed methods to tackle the performance degradation introduced by data heterogeneity, followed by empirical evaluations on challenging data distributions, highlighting strengths and possible weaknesses of the considered solutions. Finally, this Thesis presents a novel perspective on the usage of Knowledge Distillation as a mean for optimizing decentralized learning systems in settings characterized by data heterogeneity or system heterogeneity. Our vision on relevant future research directions close the manuscript.

[In search of] Digital biomarkers for objective pain assessment

Pain is a highly complex phenomenon involving intricate neural systems, whose interactions with other physiological mechanisms are not fully understood. Standard pain assessment methods, relying on verbal communication, often fail to provide reliable and accurate information, which poses a critical challenge in the clinical context. In the era of ubiquitous and inexpensive physiological monitoring, coupled with the advancement of artificial intelligence, these new tools appear as the natural candidates to be tested to address such a challenge. This thesis aims to conduct experimental research to develop digital biomarkers for pain assessment. After providing an overview of the state-of-the-art regarding pain neurophysiology and assessment tools, methods for appropriately conditioning physiological signals and controlling confounding factors are presented. The thesis focuses on three different pain conditions: cancer pain, chronic low back pain, and pain experienced by patients undergoing neurorehabilitation. The approach presented in this thesis has shown promise, but further studies are needed to confirm and strengthen these results. Prior to developing any models, a preliminary signal quality check is essential, along with the inclusion of personal and health information in the models to limit their confounding effects. A multimodal approach is preferred for better performance, although unimodal analysis has revealed interesting aspects of the pain experience. This approach can enrich the routine clinical pain assessment procedure by enabling pain to be monitored when and where it is actually experienced, and without the involvement of explicit communication,. This would improve the characterization of the pain experience, aid in antalgic therapy personalization, and bring timely relief, with the ultimate goal of improving the quality of life of patients suffering from pain.

Anatomical biomarkers of body composition in oncological population: application and development of artificial intelligences

In the Era of precision medicine and big medical data sharing, it is necessary to solve the work-flow of digital radiological big data in a productive and effective way. In particular, nowadays, it is possible to extract information “hidden” in digital images, in order to create diagnostic algorithms helping clinicians to set up more personalized therapies, which are in particular targets of modern oncological medicine. Digital images generated by the patient have a “texture” structure that is not visible but encrypted; it is “hidden” because it cannot be recognized by sight alone. Thanks to artificial intelligence, pre- and post-processing software and generation of mathematical calculation algorithms, we could perform a classification based on non-visible data contained in radiological images. Being able to calculate the volume of tissue body composition could lead to creating clasterized classes of patients inserted in standard morphological reference tables, based on human anatomy distinguished by gender and age, and maybe in future also by race. Furthermore, the branch of “morpho-radiology" is a useful modality to solve problems regarding personalized therapies, which is particularly needed in the oncological field. Actually oncological therapies are no longer based on generic drugs but on target personalized therapy. The lack of gender and age therapies table could be filled thanks to morpho-radiology data analysis application.

Meaningful insights: explainability techniques for black-box models on tabular data

Artificial Intelligence (AI) and Machine Learning (ML) are novel data analysis techniques providing very accurate prediction results. They are widely adopted in a variety of industries to improve efficiency and decision-making, but they are also being used to develop intelligent systems. Their success grounds upon complex mathematical models, whose decisions and rationale are usually difficult to comprehend for human users to the point of being dubbed as black-boxes. This is particularly relevant in sensitive and highly regulated domains. To mitigate and possibly solve this issue, the Explainable AI (XAI) field became prominent in recent years. XAI consists of models and techniques to enable understanding of the intricated patterns discovered by black-box models. In this thesis, we consider model-agnostic XAI techniques, which can be applied to Tabular data, with a particular focus on the Credit Scoring domain. Special attention is dedicated to the LIME framework, for which we propose several modifications to the vanilla algorithm, in particular: a pair of complementary Stability Indices that accurately measure LIME stability, and the OptiLIME policy which helps the practitioner finding the proper balance among explanations' stability and reliability. We subsequently put forward GLEAMS a model-agnostic surrogate interpretable model which requires to be trained only once, while providing both Local and Global explanations of the black-box model. GLEAMS produces feature attributions and what-if scenarios, from both dataset and model perspective. Eventually, we argue that synthetic data are an emerging trend in AI, being more and more used to train complex models instead of original data. To be able to explain the outcomes of such models, we must guarantee that synthetic data are reliable enough to be able to translate their explanations to real-world individuals. To this end we propose DAISYnt, a suite of tests to measure synthetic tabular data quality and privacy.

Advancing digital healthcare research and machine learning-aided knowledge extraction amidst the COVID-19 pandemic

The rapid progression of biomedical research coupled with the explosion of scientific literature has generated an exigent need for efficient and reliable systems of knowledge extraction. This dissertation contends with this challenge through a concentrated investigation of digital health, Artificial Intelligence, and specifically Machine Learning and Natural Language Processing's (NLP) potential to expedite systematic literature reviews and refine the knowledge extraction process. The surge of COVID-19 complicated the efforts of scientists, policymakers, and medical professionals in identifying pertinent articles and assessing their scientific validity. This thesis presents a substantial solution in the form of the COKE Project, an initiative that interlaces machine reading with the rigorous protocols of Evidence-Based Medicine to streamline knowledge extraction. In the framework of the COKE (“COVID-19 Knowledge Extraction framework for next-generation discovery science”) Project, this thesis aims to underscore the capacity of machine reading to create knowledge graphs from scientific texts. The project is remarkable for its innovative use of NLP techniques such as a BERT + bi-LSTM language model. This combination is employed to detect and categorize elements within medical abstracts, thereby enhancing the systematic literature review process. The COKE project's outcomes show that NLP, when used in a judiciously structured manner, can significantly reduce the time and effort required to produce medical guidelines. These findings are particularly salient during times of medical emergency, like the COVID-19 pandemic, when quick and accurate research results are critical.

Mind the gaps: cognitive-inspired AI for high-level visual sensemaking. Towards abstract concept image classification

The abundance of visual data and the push for robust AI are driving the need for automated visual sensemaking. Computer Vision (CV) faces growing demand for models that can discern not only what images "represent," but also what they "evoke." This is a demand for tools mimicking human perception at a high semantic level, categorizing images based on concepts like freedom, danger, or safety. However, automating this process is challenging due to entropy, scarcity, subjectivity, and ethical considerations. These challenges not only impact performance but also underscore the critical need for interoperability. This dissertation focuses on abstract concept-based (AC) image classification, guided by three technical principles: situated grounding, performance enhancement, and interpretability. We introduce ART-stract, a novel dataset of cultural images annotated with ACs, serving as the foundation for a series of experiments across four key domains: assessing the effectiveness of the end-to-end DL paradigm, exploring cognitive-inspired semantic intermediaries, incorporating cultural and commonsense aspects, and neuro-symbolic integration of sensory-perceptual data with cognitive-based knowledge. Our results demonstrate that integrating CV approaches with semantic technologies yields methods that surpass the current state of the art in AC image classification, outperforming the end-to-end deep vision paradigm. The results emphasize the role semantic technologies can play in developing both effective and interpretable systems, through the capturing, situating, and reasoning over knowledge related to visual data. Furthermore, this dissertation explores the complex interplay between technical and socio-technical factors. By merging technical expertise with an understanding of human and societal aspects, we advocate for responsible labeling and training practices in visual media. These insights and techniques not only advance efforts in CV and explainable artificial intelligence but also propel us toward an era of AI development that harmonizes technical prowess with deep awareness of its human and societal implications.

Intermediate linguistic task fine-tuning on Multi-lingual models

State-of-the-art NLP systems are generally based on the assumption that the underlying models are provided with vast datasets to train on. However, especially when working in multi-lingual contexts, datasets are often scarce, thus more research should be carried out in this field. This thesis investigates the benefits of introducing an additional training step when fine-tuning NLP models, named Intermediate Training, which could be exploited to augment the data used for the training phase. The Intermediate Training step is applied by training models on NLP tasks that are not strictly related to the target task, aiming to verify if the models are able to leverage the learned knowledge of such tasks. Furthermore, in order to better analyze the synergies between different categories of NLP tasks, experimentations have been extended also to Multi-Task Training, in which the model is trained on multiple tasks at the same time.

An Embedded In-Cabin Lightweight High-Performance 3D Gaze Estimation System

Gaze estimation has gained interest in recent years for being an important cue to obtain information about the internal cognitive state of humans. Regardless of whether it is the 3D gaze vector or the point of gaze (PoG), gaze estimation has been applied in various fields, such as: human robot interaction, augmented reality, medicine, aviation and automotive. In the latter field, as part of Advanced Driver-Assistance Systems (ADAS), it allows the development of cutting-edge systems capable of mitigating road accidents by monitoring driver distraction. Gaze estimation can be also used to enhance the driving experience, for instance, autonomous driving. It also can improve comfort with augmented reality components capable of being commanded by the driver's eyes. Although, several high-performance real-time inference works already exist, just a few are capable of working with only a RGB camera on computationally constrained devices, such as a microcontroller. This work aims to develop a low-cost, efficient and high-performance embedded system capable of estimating the driver's gaze using deep learning and a RGB camera. The proposed system has achieved near-SOTA performances with about 90% less memory footprint. The capabilities to generalize in unseen environments have been evaluated through a live demonstration, where high performance and near real-time inference were obtained using a webcam and a Raspberry Pi4.

Motor Unit Spike Trains Recontruction from sEMG Signals for Gesture Classification on a Low-Power Processing Platform

Hand gesture recognition based on surface electromyography (sEMG) signals is a promising approach for the development of intuitive human-machine interfaces (HMIs) in domains such as robotics and prosthetics. The sEMG signal arises from the muscles' electrical activity, and can thus be used to recognize hand gestures. The decoding from sEMG signals to actual control signals is non-trivial; typically, control systems map sEMG patterns into a set of gestures using machine learning, failing to incorporate any physiological insight. This master thesis aims at developing a bio-inspired hand gesture recognition system based on neuromuscular spike extraction rather than on simple pattern recognition. The system relies on a decomposition algorithm based on independent component analysis (ICA) that decomposes the sEMG signal into its constituent motor unit spike trains, which are then forwarded to a machine learning classifier. Since ICA does not guarantee a consistent motor unit ordering across different sessions, 3 approaches are proposed: 2 ordering criteria based on firing rate and negative entropy, and a re-calibration approach that allows the decomposition model to retain information about previous sessions. Using a multilayer perceptron (MLP), the latter approach results in an accuracy up to 99.4% in a 1-subject, 1-degree of freedom scenario. Afterwards, the decomposition and classification pipeline for inference is parallelized and profiled on the PULP platform, achieving a latency < 50 ms and an energy consumption < 1 mJ. Both the classification models tested (a support vector machine and a lightweight MLP) yielded an accuracy > 92% in a 1-subject, 5-classes (4 gestures and rest) scenario. These results prove that the proposed system is suitable for real-time execution on embedded platforms and also capable of matching the accuracy of state-of-the-art approaches, while also giving some physiological insight on the neuromuscular spikes underlying the sEMG.