Skip to main content
. 2022 Sep 24;26(12):6871–6879. doi: 10.1007/s00784-022-04724-2

Fig. 1.

Fig. 1

Typically, for (1) an input image, (2) the face is first detected and fed to (3–6) a deep learning model that (3) extracts facial features and (4) assesses the relationships between them and annotated data to (5) regress to a range of values that it is finally (6) mapped to a predicted label output