OnLIne CrOWd forecasting with data driven models – OLICOW

OLICOW aims to significantly improve capabilities to analyse and predict crowd behaviour in public areas. Crowd observation and simulation problems have been historically addressed separately by different scientific communities, and attempts to couple them are recent, have been performed offline so far, and are not yet operational for accurate real-time analysis and prediction of realistic crowd systems. OLICOW aims at designing a new generation of data-driven crowd simulation models, opening new capabilities for assimilating crowd tracking data in real-time. Besides cameras, we will rely on additional novel sensing technologies and we will
investigate up to what extent personal privacy, in the strict legal sense, may be upheld when faced with algorithms able to exploit complex correlations in multi-modal data. Our progress will allow for a more effective use of new families of sensors and facilitate crowd studies, which raise nowadays increasing questioning from the general public.