-
Mobile Localization Techniques Oriented to Tangible Web
Authors:
Osvaldo Gervasi,
Martina Fortunelli,
Riccardo Magni,
Damiano Perri,
Marco Simonetti
Abstract:
We implemented a system able to locate people indoor, with the purpose of providing assistive services. Such approach is particularly important for the Art, for providing information on exhibitions, art galleries and museums, and to allow the access to the cultural heritage patrimony to people with disabilities. The system may provide also very important information and input to elderly people, he…
▽ More
We implemented a system able to locate people indoor, with the purpose of providing assistive services. Such approach is particularly important for the Art, for providing information on exhibitions, art galleries and museums, and to allow the access to the cultural heritage patrimony to people with disabilities. The system may provide also very important information and input to elderly people, hel** them to perceive more deeply the reality and the beauty of art. The system is based on Beacons, very small and low power consumption devices, and Human Body Communication protocols. The Beacons, Bluetooth Low Energy devices, allow to obtain a position information related to predetermined reference points, and through proximity algorithms, locate a person or an object of interest. The position obtained has an error that depends from the interferences present in the area. The union of Beacons with Human Body Communication, a recent wireless technology that exploits the human body as a transmission channel, makes it possible to increase the accuracy of localization. The basic idea is to exploit the localization derived from Beacons to start a search for an electrical signal transmitted by the human body and to distinguish the position according to the information contained in the signal. The signal is transmitted by capacitance to the human body and revealed by a special resonant circuit (antenna) adapted to the microphone input of the mobile device.
△ Less
Submitted 12 December, 2022;
originally announced December 2022.
-
An Approach for Improving Automatic Mouth Emotion Recognition
Authors:
Giulio Biondi,
Valentina Franzoni,
Osvaldo Gervasi,
Damiano Perri
Abstract:
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. Th…
▽ More
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.
△ Less
Submitted 12 December, 2022;
originally announced December 2022.
-
Towards a learning-based performance modeling for accelerating Deep Neural Networks
Authors:
Damiano Perri,
Paolo Sylos Labini,
Osvaldo Gervasi,
Sergio Tasso,
Flavio Vella
Abstract:
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 Comput…
▽ More
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.
△ Less
Submitted 9 December, 2022;
originally announced December 2022.
-
High Performance Computing and Computational Intelligence Applications with MultiChaos Perspective
Authors:
Damiano Perri,
Osvaldo Gervasi,
Marco Simonetti,
Sergio Tasso
Abstract:
The experience of the COVID-19 pandemic, which has accelerated many chaotic processes in modern society, has highlighted in a very serious and urgent way the need to understand complex processes in order to achieve the common well-being. Modern High performance computing technologies, Quantum Computing, Computational Intelligence are shown to be extremely efficient and useful in safeguarding the f…
▽ More
The experience of the COVID-19 pandemic, which has accelerated many chaotic processes in modern society, has highlighted in a very serious and urgent way the need to understand complex processes in order to achieve the common well-being. Modern High performance computing technologies, Quantum Computing, Computational Intelligence are shown to be extremely efficient and useful in safeguarding the fate of mankind. These technologies are the state of the art of IT evolution and are fundamental to be competitive and efficient today. If a company is familiar with these techniques and technologies, will be able to deal with any unexpected and complicated scenario more efficiently and effectively. The main contribution of our work is a set of best practices and case studies that can help the researcher address computationally complex problems. We offer a range of software technologies, from high performance computing to machine learning and quantum computing, which represent today the state of the art to deal with extremely complex computational issues, driven by chaotic events and not easily predictable. In this chapter we analyse the different technologies and applications that will lead mankind to overcome this difficult moment, as well as to understand more and more deeply the profound aspects of very complex phenomena. In this environment of rising complexity, both in terms of technology, algorithms, and changing lifestyles, it is critical to emphasize the importance of achieving maximum efficiency and outcomes while protecting the integrity of everyone's personal data and respecting the human being as a whole.
△ Less
Submitted 25 November, 2022;
originally announced December 2022.
-
An example of use of Variational Methods in Quantum Machine Learning
Authors:
Marco Simonetti,
Damiano Perri,
Osvaldo Gervasi
Abstract:
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…
▽ More
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.
△ Less
Submitted 6 August, 2022;
originally announced August 2022.
-
IoT to monitor people flow in areas of public interest
Authors:
Damiano Perri,
Marco Simonetti,
Alex Bordini,
Simone Cimarelli,
Osvaldo Gervasi
Abstract:
The unexpected historical period we are living has abruptly pushed us to loosen any sort of interaction between individuals, gradually forcing us to deal with new ways to allow compliance with safety distances; indeed the present situation has demonstrated more than ever how critical it is to be able to properly organize our travel plans, put people in safe conditions, and avoid harmful circumstan…
▽ More
The unexpected historical period we are living has abruptly pushed us to loosen any sort of interaction between individuals, gradually forcing us to deal with new ways to allow compliance with safety distances; indeed the present situation has demonstrated more than ever how critical it is to be able to properly organize our travel plans, put people in safe conditions, and avoid harmful circumstances. The aim of this research is to set up a system to monitor the flow of people inside public places and facilities of interest (museums, theatres, cinemas, etc.) without collecting personal or sensitive data. Weak monitoring of people flows (i.e. monitoring without personal identification of the monitored subjects) through Internet of Things tools might be a viable solution to minimize lineups and overcrowding. Our study, which began as an experiment in the Umbria region of Italy, aims to be one of several answers to automated planning of people's flows in order to make our land more liveable. We intend to show that the Internet of Things gives almost unlimited tools and possibilities, from develo** a basic information process to implementing a true portal which enables business people to connect with interested consumers.
△ Less
Submitted 2 November, 2021;
originally announced November 2021.
-
Skin Cancer Classification using Inception Network and Transfer Learning
Authors:
Priscilla Benedetti,
Damiano Perri,
Marco Simonetti,
Osvaldo Gervasi,
Gianluca Reali,
Mauro Femminella
Abstract:
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 convolutio…
▽ More
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.
△ Less
Submitted 2 November, 2021;
originally announced November 2021.
-
A new method for binary classification of proteins with Machine Learning
Authors:
Damiano Perri,
Marco Simonetti,
Andrea Lombardi,
Noelia Faginas-Lago,
Osvaldo Gervasi
Abstract:
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 InceptionResNet…
▽ More
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.
△ Less
Submitted 2 November, 2021;
originally announced November 2021.
-
Binary classification of proteins by a Machine Learning approach
Authors:
Damiano Perri,
Marco Simonetti,
Andrea Lombardi,
Noelia Faginas-Lago,
Osvaldo Gervasi
Abstract:
In this work we present a system based on a Deep Learning approach, by using a Convolutional Neural Network, capable of classifying protein chains of amino acids based on the protein description contained in the Protein Data Bank. Each protein is fully described in its chemical-physical-geometric properties in a file in XML format. The aim of the work is to design a prototypical Deep Learning mach…
▽ More
In this work we present a system based on a Deep Learning approach, by using a Convolutional Neural Network, capable of classifying protein chains of amino acids based on the protein description contained in the Protein Data Bank. Each protein is fully described in its chemical-physical-geometric properties in a file in XML format. The aim of the work is to design a prototypical Deep Learning machinery for the collection and management of a huge amount of data and to validate it through its application to the classification of a sequences of amino acids. We envisage applying the described approach to more general classification problems in biomolecules, related to structural properties and similarities.
△ Less
Submitted 2 November, 2021;
originally announced November 2021.
-
Teaching Math with the help of Virtual Reality
Authors:
Marco Simonetti,
Damiano Perri,
Natale Amato,
Osvaldo Gervasi
Abstract:
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 a…
▽ More
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.
△ Less
Submitted 2 November, 2021;
originally announced November 2021.
-
Implementing a scalable and elastic computing environment based on Cloud Containers
Authors:
Damiano Perri,
Marco Simonetti,
Sergio Tasso,
Federico Ragni,
Osvaldo Gervasi
Abstract:
In this article we look at the potential of cloud containers and we provide some guidelines for companies and organisations that are starting to look at how to migrate their legacy infrastructure to something modern, reliable and scalable. We propose an architecture that has an excellent relationship between the cost of implementation and the benefits it can bring, based on the "Pilot Light" topol…
▽ More
In this article we look at the potential of cloud containers and we provide some guidelines for companies and organisations that are starting to look at how to migrate their legacy infrastructure to something modern, reliable and scalable. We propose an architecture that has an excellent relationship between the cost of implementation and the benefits it can bring, based on the "Pilot Light" topology. The services are reconfigured inside small docker containers and the workload is balanced using load balancers that allow horizontal autoscaling techniques to be exploited in the future. By generating additional containers and utilizing the possibilities given by load balancers, companies and network systems experts may model and calibrate infrastructures based on the projected number of users. Containers offer the opportunity to expand the infrastructure and increase processing capacity in a very short time. The proposed approach results in an easily maintainable and fault-tolerant system that could help and simplify the work in particular of small and medium-sized organisations.
△ Less
Submitted 2 November, 2021;
originally announced November 2021.
