FACIAL RECOGNIZER
With the aim of complementing ITER’s presence control systems as well as acting as a passive system to control the visits of people to ITER, a project is initiated to research and develop a facial recognition application that is sufficiently accurate and self-sufficient to be able to:
- Complement current presence systems with access for ITER staff.
- Detect possible intruders and raise alarms.
- To take a count of the capacities in the different buildings and industrial halls of the company.
For this purpose, a proprietary facial detection and recognition algorithm is being developed, which has a reliability of around 90% effectiveness
The system has been initially designed so that the facial recognition load does not fall directly on the client application, but rather on servers, thus increasing the detection speed and allowing its integration into thin clients in anticipation of its inclusion in other subsystems that do not have high computing capacity.
On the server side, its integration within the Teide HPC is proposed in order to maximize its potential and minimize the processing times required by the algorithm
Links of interest:
