Open Access Open Access  Restricted Access Subscription Access

Judgmental Feature Based Facial Expression Recognition and FER Datasets - A Comprehensive Study

Mahesh M Goyani, Narendra M Patel

Abstract


Facial expression is secondary supply of communication, which flows in parallel with verbal interaction. Because of its generality throughout people of different races, places, age and gender, facial expression plays a very important role in communication. Facial expressions are echo of mental state of the person. It has been intensively studied research area since last two decades. Expressions can be modeled using either descriptive features – coded using facial muscles, or judgmental features – coded using texture information. In this survey, we mainly focused on prominent judgmental feature based methods. Face is incredibly wealthy and complicated entity. Many approaches have been suggested to efficiently encode the facial expressions in lower dimensional space for effective recognition. Until now, most of the research was centered around recognition of frontal face expressions. Recent work is targeted on processing profile face and spontaneous expressions by treatment of multimodal fusion. In addition to features, dataset is another major aspect of pattern recognition. Multidimensional comparison of various facial expression databases is also derived in this paper. Additionally, survey presents scientific challenges touching the performance of system.

Full Text:

PDF