Media art course

The outputs of the mentioned projects were implemented using the software Open Face, Affectiva (AFFDEX algorithm). Within collaboration with Cultcode Institute of Visual Arts is developed and preferred, easy to use and robust an alternative to paid facial feature extraction services using OpenFace. This tool enables recognition facial action units, gaze estimation, detection of landmarks on the face and head pose, etc.



The project, as part of a diploma thesis at Faculty of arts in Presov, examined the facial expressions of female respondents during interviews about sensitive subjects topics. The design of the range of sensitive topics/questions reflected the issue of life changes of the female body as well as intimate ones experiences. One of the questions was aimed at a very common and popular consideration: „Would the respondent like a symbolic one gender change in the context of the popular opinion that men have it easier in life“. By analyzing facial expression recognition from approximately 5 hours of video interviews (7 female respondents), those moments that were typical for the display of affect were isolated in the face and the subsequent comparison with the content of the interviews in the character created their media portrait of experiencing uncertainty. The confrontation of the respondents with their already isolated expression, recorded on video without sound, was also interesting so that they had no idea what they were talking about and didn’t know that they were experiencing a certain kind of insecurity on the videos. In general, it can be said that their sight of themselves soothed and cheered them up. For many, the effect of mirroring was manifested and they imitated their own facial expressions. Such a method has the potential for further development in the intentions of artistic research as a social intervention.


The project was inspired by the concept created as part of the FF-UNIPO Intermedia creation course and deals with by taking an advertisement for pornographic content. Individual rings in which part of the observer’s face is located mean those moments while watching pornographic content when she was taken by an „explicit scene“. Within the possibility it is impossible to determine with certainty which facial expression corresponds to the preoccupation. The software could only be roughly learned changes her facial expression, watching the object when she feels strong interest. The number of rings reflects the level of engagement.