Development of an audio and video stream processing system for security and safety surveillance application
Location : IFSTTAR - Lille/Villeneuve d'Ascq, North of France
Start : as soons as possible (May 2017)
Keywords : real-time multi-cameras processing, camera network, GPU, people detection and tracking
The candidate will contribute to multi-sensors architecture and audio and video processing for security and safety applications. His objective is to develop an audio and video platform able to host processing algorithms for testing, evaluation and demonstration. The candidate will propose an architecture compatible with the processing of stored sequences and streams providing by a set of cameras or microphones. He/she will contribute on the synchronization problem and time stamping of each data stream more specifically for multi-sensor applications (multi-cameras, multi-microphones, cameras/microphones) and will propose all the tools for the future implementation on this new platform. He/She will validate its work by implementing several video based algorithms existing in the lab: a people detection and tracking by multi-views analysis and a people identification/re-identification function. He/She will test the implementation on indoor (buildings) or outdoor scene (streets) and embedded environment (inside train).
Diploma : Engineer, master degree or PhD degree
The candidate must have several competencies in the following topics:
This position is located in Lille-Villeneuve d'Ascq, within the Electronics and Signal processing Lab (LEOST) at The French Institute of Science and Technology for Transport, Development and Networks (IFSTTAR) - www.ifsttar.fr
The candidate will start as soon as possible with regards to the administrative procedures (May 2017).
(c) GdR 720 ISIS - CNRS - 2011-2015.