Human Face Analysis: From Identity to Emotion and Intention Recognition

Face recognition is the most natural mean of recognition by humans. At the same time, images (and videos) of human faces can be captured without the user’s awareness. The entertainment media and science fiction has greatly contributed in shaping the public view of these technologies, most of the times exaggerating the potential impact in one’s privacy. Even though face images can be acquired, in any place, with hidden cameras it is also true that face recognition technology is not dangerous per se. Rather, whenever properly deployed, it can result for the protection of the citizens and also enhance the user convenience.

Face recognition today has achieved a quite high performance rate and most of the problems hindering the use of this technology have now been solved. Faces can be analyzed and characterized on the basis of several features. Then, a face can be tagged with several properties, not only the bearer’s identity, but also his gender, approximate age and possible familiarity with others. Moreover, the analysis of the facial expression may also lead to understanding the mood, maybe the emotional state and intentions of the analyzed subject. May this lead to a ”Big Brother scenario”? Is this technology going to hinder a person’s freedom or privacy? These questions are still to be answered and mostly depend on tomorrow’s good use of this emerging technology. As for today, many scenarios can be envisaged where face recognition technologies can be fruitfully applied. Among them, the border control at airports and other ports of entry are just the most addressed in the recent past. Other applications still exist which have been overlooked and are yet worth a more extensive study and deployment from both the Academia and Industry.