Human behaviour supervision using multiple sensors

This project is the first step towards the construction of a full system that semantically describes people and their activities, by detecting different anomalies in their behavior. In this case, a fundamental component is the extraction of information from the human face to estimate visual semantic attributes (e.g., young woman wearing glasses is looking right). This research is part of two projects, SPACES-UPM, TIN2013-47630-C2-2-R, (http://www.geintra-uah.org/spaces/en) and HEIMDAL-UPM, TIN2016-75982-C2-2-R, (http://www.geintra-uah.org/heimdal/en/introduction), funded by the Spanish Ministry of Economy and Competitiveness.