The Artistʼs Complete Guide to Facial Expression

Emotion Recognition and Synthesis Based on MPEG‐4 FAPs

“…From this figure it is confirmed, for example, that lower − t − midlip (feature with index 3) that refers to lowering the middle of the upper lip is employed, because even the maximum value for this FAP is below zero. In the same way, the FAPs 10,12,13,14) are verified. Some of the aforementioned FAPs are described using a single variable.…”

“…Hints for this mapping were obtained from psychological studies [5,13,14] that refer to face formation during expression generation, as well as from experimental data provided by classic databases such as Ekman's and MediaLab's (see also Section 9.3). Table 9.2 illustrates the description of archetypal expressions and some variations of them using the MPEG-4 FAPs terminology.…”

“…Some of the aforementioned FAPs are described using a single variable. For example, the stretch − l − cornerlip and stretch − r − cornerlip are both modeled via f 14 (their values, shown in Table 9.4, are equal to the half value of feature f 14 ). Similar to Figure 9.3, Figure 9.4 illustrates feature statistics for the expression surprise.…”

“…Subsets of FAPs that are candidates to form an archetypal expression were defined by translating the face formations proposed by psychological studies [13,14,5] to FAPs. 2.…”

“…From the synthetic point of view, emotions that belong to the same category can be rendered by animating the same FAPs using different intensities. For example, the emotion group fear also contains worry and terror [14]; reducing or increasing the intensities of the relevant FAPs allows these two emotions to be synthesized or discriminated. In the case of expression profiles, this affects the range of variation of the corresponding FAPs that is appropriately translated.…”