This database captures faces with pose variation, expression variation and accessory variation. The aim is to simulate real-world settings for challenging face recognition task. The database includes faces of 32 peoploe with 20 males and 12 females.
For each person, we captured 120 face images with the following subcategories.
(a) frontal faces
(c) frontal faces with expression variation
(d) profiles with expression variation
(e) frontal faces with accessory variation
(f) profiles with accessory variation