In this talk, I will describe the research activities performed in PAVIS (Pattern Analysis & Computer Vision) department at the Italian Institute of Technology. In particular, the focus will be devoted to describe the multidisciplinary nature of the research and the envisaged applications. Computational vision and machine learning constitute the ground on which video surveillance and biomedical imaging problems are addressed. These two large areas, apparently very diverse, can be tackled by the same methodological approaches based on statistical and probabilistic reasoning, allowing to efficiently manage the inherent uncertain and noisy nature of the data, and even the lacking of sufficient data in some cases. The final goal is to actually design effective solutions to real-world problems. The talk will be organised with the goal to explain in general the areas and topics treated in PAVIS, while trying, when possible, to deepen technically and theoretically aspects of the some methods.
CSS e script comuni siti DOL - frase 9957