FacialActionCodingSystem

Facial Action Coding System (FACS) is a comprehensive framework for describing and measuring facial movements that observers can see on the face. Developed by psychologists Paul Ekman and Wallace Friesen in 1978, FACS decomposes facial activity into Action Units (AUs), each corresponding to anatomically defined muscle movements or groups. By coding combinations of AUs, researchers can describe complex expressions without inferring emotional states. FACS requires training to ensure reliability, and standardized manuals and software support consistency across coders.

Since its introduction, FACS has been updated and expanded with revised editions and extensions that address subtler movements, eye region actions, and mouth actions. It underpins a wide range of disciplines, including affective neuroscience, social psychology, animation, virtual agents, and human-computer interaction. In practice, researchers record facial behavior by identifying active AUs in each frame of a video or image sequence.

Applications of FACS span emotion research, deception detection, clinical assessment, literacy and communication studies, animation and game design, and robotics. Limitations include that FACS describes observable movements but does not specify why a person is moving muscles, and interpretation of emotion is context-dependent and culturally variable. Coding is labor-intensive and requires specialized training, though automation and machine learning approaches are increasingly used and often rely on FACS-based labels for supervision.

Overall, FACS provides a standardized language for facial movement that facilitates cross-study comparisons and the development of computational models of facial expression.