An analysis of a hybrid algorithm for face detection and emotion recognition.

Detecting and analyzing human face represent a necessity in many areas such as medicine, security, education, human-computer interaction, and much more. Face emotions recognition (FER) is one of the important aspects of face image analysis that is needed in many applications and technologies. In general, the research work in this area concentrates on enhancing the detection accuracy and decreasing the time needed for detection. In this work, the FER hybrid algorithm based on V-J (Viola-Jones) and CNN algorithms is tested, analyzed, and evaluated considering different aspects among which are the principle of work and factors involved and affecting it, the efficiency, the prospect of improving, and the possibility of exchanging the role between the V-J and CNN algorithms. [ABSTRACT FROM AUTHOR]