Reti Generative Avversarie nel contesto di servizi di rete sociale: un caso di studio

L’intelligenza artificiale è senza dubbio uno degli argomenti attualmente più in voga nel mondo dell’informatica, sempre in costante evoluzione ed espansione in nuovi settori. In questa elaborato progettuale viene combinato l’argomento sopracitato con il mondo dei social network, che ormai sono parte integrante della quotidianità di tutti. Viene infatti analizzato lo stato dell’arte attuale delle reti neurali, in particolare delle reti generative avversarie, e vengono esaminate le principali tipologie di social network. Su questa base, infatti, verrà realizzato un sistema di rete sociale completo nel quale una GAN sarà proprio la protagonista, sfruttando le più interessanti tecnologie attualmente disponibili. Il sistema sarà disponibile sia come applicativo per dispositivi mobile che come sito web e introdurrà elementi di gamification per aumentare l’interazione con l’utente.

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.

Design methods and geometric modelling of scaffolds for bone tissue engineering

Every year, thousand of surgical treatments are performed in order to fix up or completely substitute, where possible, organs or tissues affected by degenerative diseases. Patients with these kind of illnesses stay long times waiting for a donor that could replace, in a short time, the damaged organ or the tissue. The lack of biological alternates, related to conventional surgical treatments as autografts, allografts, e xenografts, led the researchers belonging to different areas to collaborate to find out innovative solutions. This research brought to a new discipline able to merge molecular biology, biomaterial, engineering, biomechanics and, recently, design and architecture knowledges. This discipline is named Tissue Engineering (TE) and it represents a step forward towards the substitutive or regenerative medicine. One of the major challenge of the TE is to design and develop, using a biomimetic approach, an artificial 3D anatomy scaffold, suitable for cells adhesion that are able to proliferate and differentiate themselves as consequence of the biological and biophysical stimulus offered by the specific tissue to be replaced. Nowadays, powerful instruments allow to perform analysis day by day more accurateand defined on patients that need more precise diagnosis and treatments.Starting from patient specific information provided by TC (Computed Tomography) microCT and MRI(Magnetic Resonance Imaging), an image-based approach can be performed in order to reconstruct the site to be replaced. With the aid of the recent Additive Manufacturing techniques that allow to print tridimensional objects with sub millimetric precision, it is now possible to practice an almost complete control of the parametrical characteristics of the scaffold: this is the way to achieve a correct cellular regeneration. In this work, we focalize the attention on a branch of TE known as Bone TE, whose the bone is main subject. Bone TE combines osteoconductive and morphological aspects of the scaffold, whose main properties are pore diameter, structure porosity and interconnectivity. The realization of the ideal values of these parameters represents the main goal of this work: here we'll a create simple and interactive biomimetic design process based on 3D CAD modeling and generative algorithmsthat provide a way to control the main properties and to create a structure morphologically similar to the cancellous bone. Two different typologies of scaffold will be compared: the first is based on Triply Periodic MinimalSurface (T.P.M.S.) whose basic crystalline geometries are nowadays used for Bone TE scaffolding; the second is based on using Voronoi's diagrams and they are more often used in the design of decorations and jewellery for their capacity to decompose and tasselate a volumetric space using an heterogeneous spatial distribution (often frequent in nature). In this work, we will show how to manipulate the main properties (pore diameter, structure porosity and interconnectivity) of the design TE oriented scaffolding using the implementation of generative algorithms: "bringing back the nature to the nature".

Latent representations for traditional music analysis and generation

We create and study a generative model for Irish traditional music based on Variational Autoencoders and analyze the learned latent space trying to find musically significant correlations in the latent codes' distributions in order to perform musical analysis on data. We train two kinds of models: one trained on a dataset of Irish folk melodies, one trained on bars extrapolated from the melodies dataset, each one in five variations of increasing size. We conduct the following experiments: we inspect the latent space of tunes and bars in relation to key, time signature, and estimated harmonic function of bars; we search for links between tunes in a particular style (i.e. "reels'") and their positioning in latent space relative to other tunes; we compute distances between embedded bars in a tune to gain insight into the model's understanding of the similarity between bars. Finally, we show and evaluate generative examples. We find that the learned latent space does not explicitly encode musical information and is thus unusable for musical analysis of data, while generative results are generally good and not strictly dependent on the musical coherence of the model's internal representation.

Generazione di dati sintetici di risonanza magnetica cerebrale

L’applicazione degli algoritmi di Intelligenza Artificiale (AI) al settore dell’imaging medico potrebbe apportare numerosi miglioramenti alla qualità delle cure erogate ai pazienti. Tuttavia, per poterla mettere a frutto si devono ancora superare alcuni limiti legati alla necessità di grandi quantità di immagini acquisite su pazienti reali, utili nell’addestramento degli stessi algoritmi. Il principale limite è costituito dalle norme che tutelano la privacy di dati sensibili, tra cui sono incluse le immagini mediche. La generazione di grandi dataset di immagini sintetiche, ottenute con algoritmi di Deep Learning (DL), sembra essere la soluzione a questi problemi.