Digital signage in combination with computer vision oers an eective platform for out-of-home advertisement which can oer accurate audience measurement data. Such intel- ligent digital signage systems can also adapt the displayed contents to the actual audience in real time and even enable simple non-contact human computer interaction. In this paper we would like to show how digital signage can be made much more eective by using computer vision technology. Computer vision methods for face detection, classification and recognition of people's behavior in front of digital displays can provide some objective data about the people (demographic data such as age and sex for example) who have observed the messages displayed. Furthermore, computer vision can be used for a non-contact way of user interaction with the displayed content.