Video Understanding from an Egocentric Perspective

Abstract

The research focuses on the design, development, and benchmarking of AI algorithms to extract semantic information from video acquired from an egocentric perspective. The aim is to build capabilities for novel wearable artificial agents to support humans in daily activities understanding the scene observed by a human which wear an always-on first person camera.

Keywords
artificial intelligence
deep learning
machine learning
ERC sector(s)
PE Physical Sciences and Engineering
Name supervisor
Giovanni Maria Farinella
E-mail
giovanni.farinella@unict.it
Name of Department/Faculty/School
Department of Mathematics and Computer Science
Name of the host University
University of Catania (UNICT)
EUNICE partner e-mail of destination Research
leonardo.mirabella@studium.unict.it
Country
Italy
Thesis level
PhD
Minimal language knowledge requisite
English C2
Thesis mode
On-site
Start date
Call deadline
Length of the research internship
Other
Financial support available (other than E+)
Yes