Osvaldo Gervasi

Giulio Biondi, Valentina Franzoni, Osvaldo Gervasi et al.

The study proposes and tests a technique for automated emotion recognition through mouth detection via Convolutional Neural Networks (CNN), meant to be applied for supporting people with health disorders with communication skills issues (e.g. muscle wasting, stroke, autism, or, more simply, pain) in order to recognize emotions and generate real-time feedback, or data feeding supporting systems. The software system starts the computation identifying if a face is present on the acquired image, then it looks for the mouth location and extracts the corresponding features. Both tasks are carried out using Haar Feature-based Classifiers, which guarantee fast execution and promising performance. If our previous works focused on visual micro-expressions for personalized training on a single user, this strategy aims to train the system also on generalized faces data sets.

Damiano Perri, Paolo Sylos Labini, Osvaldo Gervasi et al.

Emerging applications such as Deep Learning are often data-driven, thus traditional approaches based on auto-tuners are not performance effective across the wide range of inputs used in practice. In the present paper, we start an investigation of predictive models based on machine learning techniques in order to optimize Convolution Neural Networks (CNNs). As a use-case, we focus on the ARM Compute Library which provides three different implementations of the convolution operator at different numeric precision. Starting from a collation of benchmarks, we build and validate models learned by Decision Tree and naive Bayesian classifier. Preliminary experiments on Midgard-based ARM Mali GPU show that our predictive model outperforms all the convolution operators manually selected by the library.

Marco Simonetti, Damiano Perri, Osvaldo Gervasi

This paper introduces a deep learning system based on a quantum neural network for the binary classification of points of a specific geometric pattern (Two-Moons Classification problem) on a plane. We believe that the use of hybrid deep learning systems (classical + quantum) can reasonably bring benefits, not only in terms of computational acceleration but in understanding the underlying phenomena and mechanisms; that will lead to the creation of new forms of machine learning, as well as to a strong development in the world of quantum computation. The chosen dataset is based on a 2D binary classification generator, which helps test the effectiveness of specific algorithms; it is a set of 2D points forming two interspersed semicircles. It displays two disjointed data sets in a two-dimensional representation space: the features are, therefore, the individual points' two coordinates, $x_1$ and $x_2$. The intention was to produce a quantum deep neural network with the minimum number of trainable parameters capable of correctly recognising and classifying points.

Damiano Perri, Marco Simonetti, Andrea Lombardi et al.

In this work we set out to find a method to classify protein structures using a Deep Learning methodology. Our Artificial Intelligence has been trained to recognize complex biomolecule structures extrapolated from the Protein Data Bank (PDB) database and reprocessed as images; for this purpose various tests have been conducted with pre-trained Convolutional Neural Networks, such as InceptionResNetV2 or InceptionV3, in order to extract significant features from these images and correctly classify the molecule. A comparative analysis of the performances of the various networks will therefore be produced.

Marco Simonetti, Damiano Perri, Natale Amato et al.

In the present work we intend to introduce a system based on VR (Virtual Reality) for examining analytical-geometric structures that occur in the study of mathematics and physics concepts in the last high school classes. In our opinion, an immersive study environment has several advantages over traditional two-dimensional environments (such as a book or the simple screen of a PC or tablet), such as the spatial understanding of the concepts exposed, more peripheral awareness and moreover an evident decreasing in the information dispersion phenomenon. This does not mean that our pedagogical approach is a substitute for traditional pedagogical approaches, but is simply meant to be a robust support. In the first phase of our research we have tried to understand which mathematical objects and which tools to use to enhance mathematical teaching, to demonstrate that the use of VR techniques significantly increase the level of understanding of the mathematical subject investigated by the students.The system which provides for the integration of two machine levels, hardware and software, was subsequently tested by a representative sample of students who returned various food for thought through a questionnaire.

Priscilla Benedetti, Damiano Perri, Marco Simonetti et al.

Medical data classification is typically a challenging task due to imbalance between classes. In this paper, we propose an approach to classify dermatoscopic images from HAM10000 (Human Against Machine with 10000 training images) dataset, consisting of seven imbalanced types of skin lesions, with good precision and low resources requirements. Classification is done by using a pretrained convolutional neural network. We evaluate the accuracy and performance of the proposal and illustrate possible extensions.