17/Mar/2021 18:30 - 20:00
language of event
These talks are presented by professors of the Department of Information Engineering and Computer Science. The events are proposed and designed by students who attend the educational courses in the department.
UAVs, AI, and Retro-Remote Sensing
Unmanned Aerial Vehicles (UAVs) are recognized as very effective systems for collecting images from a very low altitude.
Their high flexibility permits immediate intervention and measurements according to customer’s specific needs. In this talk, we will overview their features and the recent evolution of UAV image analysis thanks to artificial intelligence (AI)-based methodologies. We will then discuss how AI has allowed to open a completely new research area, named retro-remote sensing, which aims at generating remotely sensed images from ancient text sources.
The enormous progresses in the field of Artificial Intelligence have allowed to develop advanced technologies for the analysis of images and videos.
Today, we have available smart systems able to identify objects, to analyse the movements and the interactions of more than one person in a scene, to understand the human emotions starting from the analysis of faces, etc.
However, the most surprising developments of the last few years have been the ones on to the production of video media. Thanks to the modern IA’s technologies it’s almost possible to create artificial, but extremely realistic videos, the so-called deepfakes, which portray people or objects that don’t exist in reality or that move in virtual contents.
In recent years, our research activity has focused on the new frontier of artificial vision and we have developed new models of deep learning, that allow to create artificial videos by animating static images or by manipulating actual videos. In particular, we have developed an innovation technology, the First Order Motion Model (FOMM), able to generate real-time deepfakes.
Lung Ultrasound, past present and future
The lung represents one of the last frontiers for ultrasound imaging technologies. The challenge posed by this organ is essentially due to the coexistence of both air and soft tissue in the volume of interest, which significantly complicates ultrasound propagation and thus imaging. Nevertheless, clinicians have since the 90’s used standard ultrasound imaging to inspect the lung and discovered clinically relevant correlations between specific imaging artifacts and a wide variety of lung pathologies. In this presentation, the origins of lung ultrasound will be reviewed together with the latest technical developments in the field, including pioneering quantitative methods capable to improve lung ultrasound specificity, the first applications of Artificial Intelligence in the classification and segmentation of lung ultrasound images, as well as the recent developments of this technology with respect to its application in the current COVID-19 pandemic.
From the largest radiotelescope to 6G: the reflector technology for the telecommunications
“Mirror mirror on the wall” – Have you ever wondered if a ‘game’ of mirrors can allow us discovering the unknown universe or calling from the most hidden places? Starting from the presentation of the largest radiotelescope in the world, the FAST, and from a vision on the cellular communications of the future, the 6G, a glimpse on the evolution and importance of the reflector technology in the telecommunications will be given.