Enactive Avatar is a research project initiated by Pia Tikka and running at Crucible Studio at Aalto University. The research is inspired by cinematographically human-like virtual character, here, enactive avatar of the live actress, responsive to viewer behavior in real-time. Non-verbal facial interaction is not only a crucial part of human-to-human communication that is worth research efforts in itself, but has even a range of technical applications from cinema to games and human-computer interfaces. The projecting is being produced using Unreal Engine, AgiSoft Photoscan and video manipulation softwares.
Tools: Unreal Engine, cameras, photogrammetry, 3Ds Max, Photoshop and others.
For the interaction in experiences where the participant connects within a real filmed actress and her digital self the 3d models from the actors must be realistic and believable enough to no create an uncanny feeling, unfortunatelly to remake a high-end version of an actor or actress in the traditional way takes a very long time, big teams and money. The Enactive Avatar project aims to prove that with new real-time technologies this pipeline or workflow can be made by small teams by the use of video knowledge and real-time rendering more than traditional 3d modelling and rigging.
This project is running at Crucible Studio & Virtual Cinema Lab, Aalto University. Conducted by Pia Tikka, doctoral and Finnish filmmaker, in collaboration with the Danish VFX supervisor Peter Hjorth (known from e.g. Lars von Trier’s Melancholia) and audiovisual artist and researcher Victor Pardinho.