|
|
|
| e-Pub |
Previous | 
Home
Section: New Results
Store Connect and Solitaria
Participants : Sébastien Gilabert, Minh Khue Phan Tran, François Brémond.
Store-Connect was a consortium aiming at detecting and positioning people in a supermarket. Several technologies were explore such as computing and merging trajectories obtained from the mobile phone of customers and from video cameras. In a second step, the goal was to detect all the ’stop’ events of the customers while shopping in the store.
SupICP
We have developed with the SED team, SupICP, a platform for integrating all plugins developed by STARS team. Our main contribution is the Ontology Language Plugin. With this plugin, we can use contextual and knowledge information inside scenarios designed for video event recognition. Currently, we are improving this plugin for combining the Ontology Language with Deep Learning technology towards “Action recognition based on Deep Learning and Ontology Language”.
We have also installed this software at the Institute Claude Pompidou, in order to conduct clinical trials, and to work with medical scientists.
Solitaria
The aim of this project is to combine data extracted from domestic sensors and from video cameras, and to implement this plugin into SupICP to monitor older people at home.