Impact of Facial Landmark Localization on Facial Expression Recognition

Although facial landmark localization (FLL) approaches are becoming increasingly accurate for characterizing regions, one question remains unanswered: what is the impact of these on subsequent related tasks? In this paper, focus put expression recognition (FER), where landmarks used face registration, which a common usage. Since most datasets do not allow proper measurement performance according to different difficulties (e.g., pose, expression, illumination, occlusion, motion blur), we also quantify recent in presence head pose variations and expressions. Finally, study FER conducted. We show that accuracy achieved so far optimizing conventional Euclidean distance does necessarily guarantee gain FER. To deal with issue, propose new evaluation metric FLL adapted