“The mimic game: real-time recognition and imitation of emotional facial expressions” by Stoiber, Aubault, Seguier and Breton

  • ©Nicolas Stoiber, Olivier Aubault, Renaud Seguier, and Gaspard Breton

Conference:


Type(s):


Title:

    The mimic game: real-time recognition and imitation of emotional facial expressions

Presenter(s)/Author(s):



Abstract:


    We present an interactive animated agent system that mimics the facial expressions of a person facing a webcam. Related studies have pursued the same objective from a retargeting perspective, by capturing the facial deformations of a source face to apply them on a virtual character one frame at a time. Our system differs from these approaches in that it works at the semantic level, and not the morphological one. The system first captures the current facial configuration of the user and recognizes his expressive state. The agent then autonomously produces a facial expression based on the identified state, using its own deformation and dynamic characteristics. This behavior is therefore referred to as mimicking instead of retargeting or cloning. Rigid head movements are also captured and mimicked by the agent.

References:


    1. Ekman, P. 1982. Emotion in the Human Face. Cambridge University Press, New York.
    2. Sattar, A., Aidarous, Y., Le Gallou, S., and Seguier, R. 2007. Face alignment by 2.5d active appearance model optimized by simplex. ICVS.
    3. Stoiber, N., Seguier, R., and Breton, G. 2009. Automatic design of a control interface for a synthetic face. Proceedings of the conference on Intelligent User Interfaces, 207–216.


ACM Digital Library Publication:



Overview Page: