HERMES is a consortium project which concentrates on how to extract descriptions of people behaviour from videos in a restricted discourse domain, such as pedestrians crossing inner-city roads, approaching or waiting at stops of buses and even humans in indoor worlds like an airport hall, a train station, or a lobby.
These video recordings (from different parts of Europe in order to prevent over-adaptation to local habits) allow to explore a coherent evaluation of human movements and facial expressions across a wide variation of scale.
Starting with the knowledge from the videos« information a system will be developed which would start with basic knowledge about pedestrian behaviour in the chosen discourse domain, but could cluster evaluation results into semantically meaningful subsets of behaviours.