[1] It describes a standard for virtually representing humans and humanoids in a way that adequately achieves visual speech intelligibility as well as the mood and gesture of the speaker, and allows for very low bitrate compression and transmission of animation parameters.
[2] FAPs control key feature points on a face model mesh that are used to produce animated visemes and facial expressions, as well as head and eye movement.
[1] These feature points are part of the Face Definition Parameters (FDPs) also defined in the MPEG-4 standard.
[3] FAPs represent 66 displacements and rotations of the feature points from the neutral face position, which is defined as: mouth closed, eyelids tangent to the iris, gaze and head orientation straight ahead, teeth touching, and tongue touching teeth.
[4] These FAPs were designed to be closely related to human facial muscle movements.