Visiting Tampere University

Visiting Tampere University

Lo studente di dottorato Michael Neri farà visita alla Tampere University a Tampere, Finlandia, nel periodo Gennaio-Aprile 2023. Collaborerà con l’Audio Research Group, supervisionato dal Prof. Virtanen. Il lavoro di ricerca si concentrerà sulla stima della distanza tra un microfono e un evento sonoro, sfruttando soltanto l’informazione di un solo canale audio.

 

———————————————-

PhD student Michael Neri will visit Tampere University in Tampere, Finland during January-April 2023. He will collaborate with the Audio Research Group, supervised by Professor Virtanen. The research work will focus on estimating the distance between a microphone and a sound event, using only single channel audio information.

 

Photos

Age of Information from multiple strategic sources analyzed via game theory, Prof. Leonardo Badia

Abstract

Age of Information (AoI) is defined as the time elapsed since the last correct reception of information from a source. It is a performance metric recently addressed by some authors as more important than throughput e.g. for sensing applications – in fact, sending data bursts and then not updating for a long time is worse than sparsely but regularly keeping the receiver up-to-date.

This talk investigates a system with strategic sources (meaning they use game theory), making independent and periodic decisions whether to send an update, where they balance the expected utility of updating versus the cost in doing so. Uncoordinated transmissions are often considered inefficient and unstable from a game theoretic perspective, because the players (i.e., the transmitters) want to maximize their individual utility and tend to be aggressive. We highlight that our scenario can be formalized as a game, whose Nash equilibria present some catchy topological results, also hinting at the possible design choices for achieving overall efficiency of a distributed control.

Short bio of Prof. Leonardo Badia

Leonardo Badia received both the Laurea degree (summa cum laude) in electrical engineering and the Ph.D. in information engineering (with highest mark) from the University of Ferrara, in year 2000 and 2004, respectively. From 2000 to 2006 he was with the Department of Engineering, University of Ferrara. Between 2002 and 2003 he visited the RST Labs (now Wireless@KTH) at the Royal Institute of Technology (KTH), Stockholm, Sweden. In May 2006, he moved to the IMT Lucca Institute for Advanced Studies, Lucca, Italy, as an Assistant Professor. In March 2011 he joined the faculty of the University of Padova as an Assistant Professor. In 2016, he was promoted to Associate Professor.
He is author of more than 200 scientific papers published in international peer-reviewed journals or conferences. His research interests are in the broad area of mathematical optimization for communication networks, including analysis of transmission protocols via Markov models; cross-layer optimization of routing/scheduling/resource allocation; Age-of-Information; models for energy harvesting; applications of game theory to communication networks.

Details

When: 07/12/2022

Where: Sala Multimediale @ DIIEM, Roma Tre University

Maker Faire Rome 2022

The MUSE Research Group participated in the 2022 Edition of the Maker Faire Rome!

We presented ‘I know that feel, bro!’ a system which automatically recognizes emotions from human body language and postures. Starting from the videos recorded by an RGB camera, we extract the skeleton joints and process them through an on-purpose designed Artificial Intelligence architecture. By processing only the skeleton joints we perform emotion classification without exploiting the facial expressions, thus being privacy compliant. Our system is currently able to classify four emotions: boredom, disgust, happiness and interest.

Our project is one of the four selected by Rai Scuola for the Maker Faire 2022 Special. You can find our video at https://www.raiscuola.rai.it/tecnologia/articoli/2022/10/I-know-that-feel-bro-fc6a3474-9a68-45ff-879c-1c422ada4eb6.html!!!

10th European Workshop on Visual Information Processing (EUVIP)

10th European Workshop on Visual Information Processing (EUVIP)

Alcuni Dottorandi e Professori del laboratorio COMLAB hanno partecipato alla decima edizione del l’European Workshop on Visual Information Processing (EUVIP 2022) in cui sono stati presentati tre lavori di ricerca:
– “3D Object Detection On Synthetic Point Clouds For Railway Applications”
– “ParalMGC Multiple Audio Representations for Synthetic Human Speech Attribution”
– “HEROES: a video-based Human Emotion Recognition database”

Riunione Annuale GTTI 2022

Come ogni anno si è tenuta la riunione del Gruppo Telecomunicazioni e Tecnologie dell’informazione (GTTI). Grazie a queste riunioni Dottorandi, Assegnisti, Ricercatori e Professori di diverse università italiane hanno avuto l’opportunità di interagire e confrontarsi sugli sviluppi scientifici , tecnologici didattici del settore telecomunicazioni… e su molto altro ancora!


European Researcher’s Night 2022

AI-based emotion recognition through body language: “we know that feel bro!”

Friday 30 September, 2022 was the European Researcher’s Night at Roma Tre University! Since 2005 this event has been bringing research and researchers closer to the public, with the aim of promoting excellent research projects across Europe and beyond, increasing the interest of young people in science and research careers, and showcasing the impact of researchers’ work on people’s daily lives in fun, inspiring ways. The event is part of the #LEAF (heaL thE plAnet’s Future) project, organized by Frascati Scienza and promoted by the European Commission.

For more information on the event: https://nottericerca.uniroma3.it/

We presented an automatic emotion recognition system based on the analysis of body language through Artificial Intelligence, to evaluate the Quality of Experience of a user. For instance, when a person is interested in something he/she stops in front of or gets closer to it. Therefore, the movements are slow or absent, the trunk is in an upright position or bent towards the object of interest, and the head may be tilted. When a person is bored, on the contrary, he/she does not stop, and performs sluggish and slow movements. Based on these behaviors we classify the emotion felt by the user.

For more information on the project: https://muse.uniroma3.it/heroes/

Visualizing San Saba Summerschool

Abstract:

This Summer School exploits a bottom-up strategy to teach how to employ new technologies such as laser scanners, 360 cameras, termocameras, GNSS and specific software for realizing a virtual tour of the “Basilica di San Saba” in Rome. The students had the opportunity to meet experts in the Digital Cultural Heritage field and to gain insights concerning 3D multimedia, Virtual Reality, and Augmented Reality.

The school has been organized by the Department of Humanities Studies of Roma Tre University.

Find out more at https://studiumanistici.uniroma3.it/didattica/post-lauream/visualizing-san-saba-geotecnologie-applicate-ai-beni-culturali/.

Details:

When: July 4 to July 10

Where: Basilica di San Saba, Rome

Visualizing San Saba

Ethical hacking seminar

Abstract:

Ethical hacking can be defined as the security practice in which a security expert attempts to break into a system, simulating the actions of a malicious cyber-attacker. These techniques play an essential role in checking for weaknesses and entry points in networks and infrastructures. This practice, in fact, helps the identification of security vulnerabilities which can then be resolved before a malicious attacker has the opportunity to exploit them. To achieve this goal, the security expert needs to use the same methods a malicious attacker would use. Given the value of ethical hacking, especially considering the damage caused by a successful malicious attack, there is an increasing interest in this field.

The seminar consists of two hands-on lectures in which the participants will have the opportunity to perform different types of attacks going from passive techniques, such as network sniffing, to active attacks, such as a man-in-the-middle attack aiming at password sniffing. The attacks will be performed exploiting virtual machinesto mimic the victim and attacker computers. The virtual machine configuration, as well as the theory concepts associated to the considered attacks will be discussed during the seminar. However, the virtual machine installation has to be performed before the seminar day. For this reason, a tutorial for the required installations is provided here.

Details:

When: May 26 and 27, 2022 from 8:00 to 10:00

Where: Room N2