Facial expression recognition in the spatio-temporal domain

Abstract

In recent years, automatic Facial Expression Recognition (FER) has been extensively studied and has gained much research interest due to its many useful applications. However, to the best of our knowledge, the proposed algorithms all suffer from some limitations, which make the systems work inaccurately and unstably in the real scenaros. In this thesis, we aim to propose and investigate a full automatic system to give effective and accurate classification of facial expressions.