Seminari di Informatica: tutti i giovedì presso il DIBRIS I seminari in Informatica saranno tenuti da dottorandi e docenti, nella Sala conferenze del DIBRIS.
The talks will be recorded and they will be available online at the YouTube channel Unige-DIBRIS. If you don't want to miss anything, add the seminars to your calendar following this link.
Date: Thursday, December 12th, 2019
Location: Conference Room (3rd floor) DIBRIS Via Dodecaneso 35 - Genoa (Italy)
Time: 11:30 AM
Speakers: Andrea Gaggioli and Alice Chirico (Catholic University of Milan, Italy)
Host: Stefano Rovetta
Title: Positive Technology: Promoting emotional well-being and positive change through advanced interactive technologies
Abstract: Although many scientific efforts have been devoted to acknowledging the risks of digital technologies, the question of how computers could be used to improve people's well-being hasbeen much less explored. This was the main motivation for the development of a novel research area—Positive Technology—which aims at investigating how technology-based applications and services can be used to foster positive growth of individuals, groups and institutions. In the last 10 years, research in Positive Technology has attracted increasing attention from an interdisciplinary community of scholars, leading to many conference papers, dedicated symposia and workshops, special issues in journals, and edited books. As an emerging area of research, considerable efforts have been spent on developing conceptual pillars and levels of analysis, as well as on the definition of frameworks for bringing well-being principles into the design of interactive systems. At the methodological and applied level, research on Positive Technology has focused on the design, development, and validation of novel digital experiences that aims at promoting positive change through pleasure, flow, meaning, competence, and positive relationships. In this talk, we will focus on the main tenets of positive technology, with a particular focus on the role that digital tools can play in promoting emotional wellbeing.
Short bio: Andrea Gaggioli is full professor of psychology at Università Cattolica del Sacro Cuore in Milan. His main focus is on "Positive Technology", a topic at the intersection of psychology, neuroscience and interactive design, which studies how digital tools can be used to help individuals achieve greater wellbeing. Dr Gaggioli has published over 150 articles in peer-reviewed journals, including Science and Nature and served as a keynote at EuroVR, Persuasive Technology, Design & Emotion, Supporting Health by Tech conferences and several others. His research has been featured in international media outlets, including The New York Times, Die Zeit, Reuters and Scientific American.
Alice Chirico is post-doc in psychology at the at the Department of Psychology, Catholic University of the Sacred Heart. Alice's main research focus is on complex emotional experiences, such as awe and flow, and on the role of advanced interactive technologies for the study of these affective states.
Date: Thursday, 12th December 2019
Location: Dibris, Valletta Puggia, Conference Hall (322)
Speaker: Lorenzo Benvenuto
Title: GNSS Android: a solution for improving GNSS positioning in IoT devices
The positioning part in fact is a crucial aspect for most IoT applications: the major part of IoT devices must be geolicalized in order to fulfil its function. The common way in which IoT devices are geolocalized is through GNSS (Global Navigation Satellite System) sensor. Considering IoT the GNSS devices that best match the requested features are those embedded into smartphone. Starting from May 2016 Google announced the availability of raw GNSS observation on Android devices: this means that, starting from Android 7, users can access GNSS measurement and not just the position. Starting from all these considerations, the purpose of this research is to improve the integrity and the positioning accuracy of smartphone compensating for some low hardware features (in particular the low quality of the antenna) via software. Excluding satellite affected by multipath can be a good option to improve the positioning accuracy. This is what the MDP algorithm (developed by GTER) does. It’s main function is to pre-process the signals from all the tracked satellites recognizing the onces potentially affected by multipath and excluding them from the positioning computation. So the goal of this research activity is to implement the MDP algorithm for Android devices also proposing some modify or improvement to the algorithm itself on the basis of the results of future test that will be done. This seminar will initially present the basics of GNSS positioning, paying particular attention to raw data obtained from ANDROID devices. Subsequently, the first positioning tests carried out with the xiaomi mi 9 device will be illustrated: the problems encountered, the positioning performance compared to other receivers, and the first steps for the implementation of the MDP algorithm on ANDROID devices will be illustrated. Short bio:
Lorenzo got his master’s degree in civil and environmental engineering (curriculum “Engineering for the environment and the territory) in July 2018 at the Polytechnic School of the University of Genoa. During his studies he particularly focused on geomatics and GNSS developing a bachelor’s thesis on low cost GNSS receiver for precise positioning, and a master’s thesis on the research field “GNSS Meteorology”. Nowadays he continues his studies attending the XXXIV cycle of the PhD in Computer Science and Systems Engineering at University of Genoa (DIBRIS), in collaboration with Gter srl. The main purpose of his research is to realise an IoT device for precise satellite positioning, composed by low-cost GNSS receiver (such as the smartphone’s once).
Date: Thursday, December 19th, 2019
Location: Conference Room (3rd floor) DIBRIS Via Dodecaneso 35 - Genoa (Italy)
Time: 3:30 PM
Speaker: Jianyi Lin (kalyfa University, Abu Dhabi, UAE)
Title: Sparsity models and algorithms for biosignal processing and computer vision
Abstract: Based on Occam's Razor principle, the sparse representation model is concerned with finding the solution of an underdetermined linear system having the minimal number of non-null components. The computational difficulties arising in tackling this non-convex combinatorial problem motivated the investigations on structural properties to ease the conception of effective algorithms for approximating the optimal solution, such as Matching Pursuit, Basis Pursuit or Lasso. I will present a recently devised sparse decomposition technique based on the fixed-point iteration of Lipschitz-continuous mappings over the set of feasible solutions, as well as the evidence of its convergence properties and performances in Phase Transitions. As first application of this sparsity technique, we will see that it can be blended into the process of succinctly encoding Electrocardiogram signals using a natural basis representation of heart-beat segments, resulting in a highly compressive encoding scheme even for abnormal heart records. The devised techniques has been also applied in computer vision to the so-called Single Sample Per Person (SSPP) Face Recognition systems. Those systems hardly benefit from Deep Learning to the full extent, owing to the lack of sample images, as typical in small sample-size problems. Hence, the related SSPP problems have been receiving an increasing attention due to the challenges posed when conceiving real-world applications in uncontrolled or wild environments.
Short bio: Jianyi Lin is assistant professor at the Department of Mathematics, Khalifa University of Science and Technology, Abu Dhabi, UAE. After completing his PhD in Mathematics and Statistics for Computational Sciences at the University of Milan, he worked for five years as a postdoctoral fellow. He has previous research and teaching experience at Politecnico di Milano and University of Milan, for which he is still an affiliated member of the Perceptual Computing and Human Sensing Lab. During his research activity, he visited the Birkbeck Institute of Data Analytics, University of London and the University of Zurich, he collaborated in 2 basic research projects and 1 industrial research project, and he is Co-PI of two applied research grants. His main research topics span computational complexity of geometric clustering, sparse signal modelling, computational methods for bioinformatics and stochastic models for pattern statistics.
Hoping to see you there!