Project CAVIAR CApturing VIewers' Affective Response.

In this project we propose the automatic analysis of the relation between the audiovisual characteristics of a multimedia production and the impact caused in its audience. With this aim, potential synergies are explored between different areas of knowledge including, among others: audiovisual communication, computer vision, multimodal systems, biometric sensors, social network analysis, opinion mining, and affective computing. Our efforts are oriented towards combining these technologies to introduce novel computational models that could predict the reactions of spectators to multimedia elements across different media and moments. On the one hand, we study the cognitive and emotional response of the spectators while they are watching the media instances, using neuroscience techniques and biometric sensors. On the other hand, we also study the reaction shown by the audience on social networks by relying on the automatic collection and analysis of different metadata related to the media elements, such as popularity, sharing patterns, ratings and commentaries.