Equipped with cameras, the participants will be wandering around Chaumont to collect any typographic signs they cross, regardless of their type : ads, posters, signs, road signs or graffiti for intance. After this walk, the photos will be processed by specific algorithms ( OCR optical characters recognition) in order to automatically extract anything the machine will have been able to decode, by also associating other more contextual information such as position, dominating colours and neighbouring objects. The participants will then think about how to use these computational resources as raw material for imagining and producing a graphic composition testifying their personal wandering.