Deep learning for computer vision: a comparison between convolutional neural networks and hierarchical temporal memories on object recognition tasks

In recent years, Deep Learning techniques have shown to perform well on a large variety of problems both in Computer Vision and Natural Language Processing, reaching and often surpassing the state of the art on many tasks. The rise of deep learning is also revolutionizing the entire field of Machine Learning and Pattern Recognition pushing forward the concepts of automatic feature extraction and unsupervised learning in general. However, despite the strong success both in science and business, deep learning has its own limitations. It is often questioned if such techniques are only some kind of brute-force statistical approaches and if they can only work in the context of High Performance Computing with tons of data. Another important question is whether they are really biologically inspired, as claimed in certain cases, and if they can scale well in terms of "intelligence". The dissertation is focused on trying to answer these key questions in the context of Computer Vision and, in particular, Object Recognition, a task that has been heavily revolutionized by recent advances in the field. Practically speaking, these answers are based on an exhaustive comparison between two, very different, deep learning techniques on the aforementioned task: Convolutional Neural Network (CNN) and Hierarchical Temporal memory (HTM). They stand for two different approaches and points of view within the big hat of deep learning and are the best choices to understand and point out strengths and weaknesses of each of them. CNN is considered one of the most classic and powerful supervised methods used today in machine learning and pattern recognition, especially in object recognition. CNNs are well received and accepted by the scientific community and are already deployed in large corporation like Google and Facebook for solving face recognition and image auto-tagging problems. HTM, on the other hand, is known as a new emerging paradigm and a new meanly-unsupervised method, that is more biologically inspired. It tries to gain more insights from the computational neuroscience community in order to incorporate concepts like time, context and attention during the learning process which are typical of the human brain. In the end, the thesis is supposed to prove that in certain cases, with a lower quantity of data, HTM can outperform CNN.

Machine learning and convolutional neural networks for melanomas identification

Il machine learning negli ultimi anni ha acquisito una crescente popolarità nell’ambito della ricerca scientifica e delle sue applicazioni. Lo scopo di questa tesi è stato quello di studiare il machine learning nei suoi aspetti generali e applicarlo a problemi di computer vision. La tesi ha affrontato le difficoltà del dover spiegare dal punto di vista teorico gli algoritmi alla base delle reti neurali convoluzionali e ha successivamente trattato due problemi concreti di riconoscimento immagini: il dataset MNIST (immagini di cifre scritte a mano) e un dataset che sarà chiamato ”MELANOMA dataset” (immagini di melanomi e nevi sani). Utilizzando le tecniche spiegate nella sezione teorica si sono riusciti ad ottenere risultati soddifacenti per entrambi i dataset ottenendo una precisione del 98% per il MNIST e del 76.8% per il MELANOMA dataset

Generative Neural Networks for image iper-resolution and improvement of Optical Character Recognition's performances

The usage of Optical Character Recognition’s (OCR, systems is a widely spread technology into the world of Computer Vision and Machine Learning. It is a topic that interest many field, for example the automotive, where becomes a specialized task known as License Plate Recognition, useful for many application from the automation of toll road to intelligent payments. However, OCR systems need to be very accurate and generalizable in order to be able to extract the text of license plates under high variable conditions, from the type of camera used for acquisition to light changes. Such variables compromise the quality of digitalized real scenes causing the presence of noise and degradation of various type, which can be minimized with the application of modern approaches for image iper resolution and noise reduction. Oneclass of them is known as Generative Neural Networks, which are very strong ally for the solution of this popular problem.

Fluctuation properties in random walks on networks and simple integrate and fire models

In questa tesi si è studiato l’insorgere di eventi critici in un semplice modello neurale del tipo Integrate and Fire, basato su processi dinamici stocastici markoviani definiti su una rete. Il segnale neurale elettrico è stato modellato da un flusso di particelle. Si è concentrata l’attenzione sulla fase transiente del sistema, cercando di identificare fenomeni simili alla sincronizzazione neurale, la quale può essere considerata un evento critico. Sono state studiate reti particolarmente semplici, trovando che il modello proposto ha la capacità di produrre effetti "a cascata" nell’attività neurale, dovuti a Self Organized Criticality (auto organizzazione del sistema in stati instabili); questi effetti non vengono invece osservati in Random Walks sulle stesse reti. Si è visto che un piccolo stimolo random è capace di generare nell’attività della rete delle fluttuazioni notevoli, in particolar modo se il sistema si trova in una fase al limite dell’equilibrio. I picchi di attività così rilevati sono stati interpretati come valanghe di segnale neurale, fenomeno riconducibile alla sincronizzazione.

Exact Combinatorial Optimization with Graph Convolutional Neural Networks

Combinatorial optimization problems are typically tackled by the branch-and-bound paradigm. We propose to learn a variable selection policy for branch-and-bound in mixed-integer linear programming, by imitation learning on a diversified variant of the strong branching expert rule. We encode states as bipartite graphs and parameterize the policy as a graph convolutional neural network. Experiments on a series of synthetic problems demonstrate that our approach produces policies that can improve upon expert-designed branching rules on large problems, and generalize to instances significantly larger than seen during training.

Neural Network Based UAV Fault Tolerant Control with Aerodynamic Disturbance Estimation and Compensation

In this thesis, the problem of controlling a quadrotor UAV is considered. It is done by presenting an original control system, designed as a combination of Neural Networks and Disturbance Observer, using a composite learning approach for a system of the second order, which is a novel methodology in literature. After a brief introduction about the quadrotors, the concepts needed to understand the controller are presented, such as the main notions of advanced control, the basic structure and design of a Neural Network, the modeling of a quadrotor and its dynamics. The full simulator, developed on the MATLAB Simulink environment, used throughout the whole thesis, is also shown. For the guidance and control purposes, a Sliding Mode Controller, used as a reference, it is firstly introduced, and its theory and implementation on the simulator are illustrated. Finally the original controller is introduced, through its novel formulation, and implementation on the model. The effectiveness and robustness of the two controllers are then proven by extensive simulations in all different conditions of external disturbance and faults.

Predicting corticospinal excitability from pre-TMS EEG using convolutional neural networks

The amplitude of motor evoked potentials (MEPs) elicited by transcranial magnetic stimulation (TMS) of the primary motor cortex (M1) shows a large variability from trial to trial, although MEPs are evoked by the same repeated stimulus. A multitude of factors is believed to influence MEP amplitudes, such as cortical, spinal and motor excitability state. The goal of this work is to explore to which degree the variation in MEP amplitudes can be explained by the cortical state right before the stimulation. Specifically, we analyzed a dataset acquired on eleven healthy subjects comprising, for each subject, 840 single TMS pulses applied to the left M1 during acquisition of electroencephalography (EEG) and electromyography (EMG). An interpretable convolutional neural network, named SincEEGNet, was utilized to discriminate between low- and high-corticospinal excitability trials, defined according to the MEP amplitude, using in input the pre-TMS EEG. This data-driven approach enabled considering multiple brain locations and frequency bands without any a priori selection. Post-hoc interpretation techniques were adopted to enhance interpretation by identifying the more relevant EEG features for the classification. Results show that individualized classifiers successfully discriminated between low and high M1 excitability states in all participants. Outcomes of the interpretation methods suggest the importance of the electrodes situated over the TMS stimulation site, as well as the relevance of the temporal samples of the input EEG closer to the stimulation time. This novel decoding method allows causal investigation of the cortical excitability state, which may be relevant for personalizing and increasing the efficacy of therapeutic brain-state dependent brain stimulation (for example in patients affected by Parkinson’s disease).

Tree regularized neural networks for key point analysis

This thesis contributes to the ArgMining 2021 shared task on Key Point Analysis. Key Point Analysis entails extracting and calculating the prevalence of a concise list of the most prominent talking points, from an input corpus. These talking points are usually referred to as key points. Key point analysis is divided into two subtasks: Key Point Matching, which involves assigning a matching score to each key point/argument pair, and Key Point Generation, which consists of the generation of key points. The task of Key Point Matching was approached using different models: a pretrained Sentence Transformers model and a tree-constrained Graph Neural Network were tested. The best model was the fine-tuned Sentence Transformers, which achieved a mean Average Precision score of 0.75, ranking 12 compared to other participating teams. The model was then used for the subtask of Key Point Generation using the extractive method in the selection of key point candidates and the model developed for the previous subtask to evaluate them.

Segmentazione di immagini mammografiche con convolutional neural networks

Il tumore al seno si colloca al primo posto per livello di mortalità tra le patologie tumorali che colpiscono la popolazione femminile mondiale. Diversi studi clinici hanno dimostrato come la diagnosi da parte del radiologo possa essere aiutata e migliorata dai sistemi di Computer Aided Detection (CAD). A causa della grande variabilità di forma e dimensioni delle masse tumorali e della somiglianza di queste con i tessuti che le ospitano, la loro ricerca automatizzata è un problema estremamente complicato. Un sistema di CAD è generalmente composto da due livelli di classificazione: la detection, responsabile dell’individuazione delle regioni sospette presenti sul mammogramma (ROI) e quindi dell’eliminazione preventiva delle zone non a rischio; la classificazione vera e propria (classification) delle ROI in masse e tessuto sano. Lo scopo principale di questa tesi è lo studio di nuove metodologie di detection che possano migliorare le prestazioni ottenute con le tecniche tradizionali. Si considera la detection come un problema di apprendimento supervisionato e lo si affronta mediante le Convolutional Neural Networks (CNN), un algoritmo appartenente al deep learning, nuova branca del machine learning. Le CNN si ispirano alle scoperte di Hubel e Wiesel riguardanti due tipi base di cellule identificate nella corteccia visiva dei gatti: le cellule semplici (S), che rispondono a stimoli simili ai bordi, e le cellule complesse (C) che sono localmente invarianti all’esatta posizione dello stimolo. In analogia con la corteccia visiva, le CNN utilizzano un’architettura profonda caratterizzata da strati che eseguono sulle immagini, alternativamente, operazioni di convoluzione e subsampling. Le CNN, che hanno un input bidimensionale, vengono solitamente usate per problemi di classificazione e riconoscimento automatico di immagini quali oggetti, facce e loghi o per l’analisi di documenti.

Integrating wireless sensor networks and internet of things: A coap-based approach

L'obiettivo su cui è stata basata questa Tesi di Laurea è stato quello di integrare la tecnologia delle Wireless Sensor Networks (WSN) al contesto dell'Internet delle cose (IoT). Per poter raggiungere questo obiettivo, il primo passo è stato quello di approfondire il concetto dell'Internet delle cose, in modo tale da comprendere se effettivamente fosse stato possibile applicarlo anche alle WSNs. Quindi è stata analizzata l'architettura delle WSNs e successivamente è stata fatta una ricerca per capire quali fossero stati i vari tipi di sistemi operativi e protocolli di comunicazione supportati da queste reti. Infine sono state studiate alcune IoT software platforms. Il secondo passo è stato quindi di implementare uno stack software che abilitasse la comunicazione tra WSNs e una IoT platform. Come protocollo applicativo da utilizzare per la comunicazione con le WSNs è stato usato CoAP. Lo sviluppo di questo stack ha consentito di estendere la piattaforma SensibleThings e il linguaggio di programmazione utilizzato è stato Java. Come terzo passo è stata effettuata una ricerca per comprendere a quale scenario di applicazione reale, lo stack software progettato potesse essere applicato. Successivamente, al fine di testare il corretto funzionamento dello stack CoAP, è stata sviluppata una proof of concept application che simulasse un sistema per la rilevazione di incendi. Questo scenario era caratterizzato da due WSNs che inviavano la temperatura rilevata da sensori termici ad un terzo nodo che fungeva da control center, il cui compito era quello di capire se i valori ricevuti erano al di sopra di una certa soglia e quindi attivare un allarme. Infine, l'ultimo passo di questo lavoro di tesi è stato quello di valutare le performance del sistema sviluppato. I parametri usati per effettuare queste valutazioni sono stati: tempi di durata delle richieste CoAP, overhead introdotto dallo stack CoAP alla piattaforma Sensible Things e la scalabilità di un particolare componente dello stack. I risultati di questi test hanno mostrato che la soluzione sviluppata in questa tesi ha introdotto un overheadmolto limitato alla piattaforma preesistente e inoltre che non tutte le richieste hanno la stessa durata, in quanto essa dipende dal tipo della richiesta inviata verso una WSN. Tuttavia, le performance del sistema potrebbero essere ulteriormente migliorate, ad esempio sviluppando un algoritmo che consenta la gestione concorrente di richieste CoAP multiple inviate da uno stesso nodo. Inoltre, poichè in questo lavoro di tesi non è stato considerato il problema della sicurezza, una possibile estensione al lavoro svolto potrebbe essere quello di implementare delle politiche per una comunicazione sicura tra Sensible Things e le WSNs.

Mathematical models for cellular aggregation: the chemotactic instability and clustering formation

In this thesis we present a mathematical formulation of the interaction between microorganisms such as bacteria or amoebae and chemicals, often produced by the organisms themselves. This interaction is called chemotaxis and leads to cellular aggregation. We derive some models to describe chemotaxis. The first is the pioneristic Keller-Segel parabolic-parabolic model and it is derived by two different frameworks: a macroscopic perspective and a microscopic perspective, in which we start with a stochastic differential equation and we perform a mean-field approximation. This parabolic model may be generalized by the introduction of a degenerate diffusion parameter, which depends on the density itself via a power law. Then we derive a model for chemotaxis based on Cattaneo's law of heat propagation with finite speed, which is a hyperbolic model. The last model proposed here is a hydrodynamic model, which takes into account the inertia of the system by a friction force. In the limit of strong friction, the model reduces to the parabolic model, whereas in the limit of weak friction, we recover a hyperbolic model. Finally, we analyze the instability condition, which is the condition that leads to aggregation, and we describe the different kinds of aggregates we may obtain: the parabolic models lead to clusters or peaks whereas the hyperbolic models lead to the formation of network patterns or filaments. Moreover, we discuss the analogy between bacterial colonies and self gravitating systems by comparing the chemotactic collapse and the gravitational collapse (Jeans instability).

SynBA: A contextualized Synonim-Based adversarial Attack for text classification

With the advent of high-performance computing devices, deep neural networks have gained a lot of popularity in solving many Natural Language Processing tasks. However, they are also vulnerable to adversarial attacks, which are able to modify the input text in order to mislead the target model. Adversarial attacks are a serious threat to the security of deep neural networks, and they can be used to craft adversarial examples that steer the model towards a wrong decision. In this dissertation, we propose SynBA, a novel contextualized synonym-based adversarial attack for text classification. SynBA is based on the idea of replacing words in the input text with their synonyms, which are selected according to the context of the sentence. We show that SynBA successfully generates adversarial examples that are able to fool the target model with a high success rate. We demonstrate three advantages of this proposed approach: (1) effective - it outperforms state-of-the-art attacks by semantic similarity and perturbation rate, (2) utility-preserving - it preserves semantic content, grammaticality, and correct types classified by humans, and (3) efficient - it performs attacks faster than other methods.

Framework for anomaly detection on photovoltaic plants

Technological advancement has undergone exponential growth in recent years, and this has brought significant improvements in the computational capabilities of computers, which can now perform an enormous amount of calculations per second. Taking advantage of these improvements has made it possible to devise algorithms that are very demanding in terms of the computational resources needed to develop architectures capable of solving the most complex problems: currently the most powerful of these are neural networks and in this thesis I will combine these tecniques with classical computer vision algorithms to improve the speed and accuracy of maintenance in photovoltaic facilities.

Edge-aware Graph Attention Networks: Joint Reasoning On Text and Knowledge Graphs for Biomedical Question Answering

Much of the real-world dataset, including textual data, can be represented using graph structures. The use of graphs to represent textual data has many advantages, mainly related to maintaining a more significant amount of information, such as the relationships between words and their types. In recent years, many neural network architectures have been proposed to deal with tasks on graphs. Many of them consider only node features, ignoring or not giving the proper relevance to relationships between them. However, in many node classification tasks, they play a fundamental role. This thesis aims to analyze the main GNNs, evaluate their advantages and disadvantages, propose an innovative solution considered as an extension of GAT, and apply them to a case study in the biomedical field. We propose the reference GNNs, implemented with methodologies later analyzed, and then applied to a question answering system in the biomedical field as a replacement for the pre-existing GNN. We attempt to obtain better results by using models that can accept as input both node and edge features. As shown later, our proposed models can beat the original solution and define the state-of-the-art for the task under analysis.