Basic face emotion detector

Title: Recognizing and classifying emotions on the human face with convolutional neural networks

Abstract proposal:

Develop an OpenCV-Python-based system that is capable of recognizing the facial gestures, the so-called six basic human emotions:

• joy
• sorrow
• anger
• fear
• surprise
• disgust

based on the movements and the position of the human face. The application understands the fine and complex changes of the faces. Examine methods of associating the key points and movements of the face with emotions. Identify the various segments of the face:

• stems
• wrinkles
• eyes
• eyebrows
• mouth

and then work on the stand-alone and relative position of movement of these key areas. Based on this study, based on various previous researches (eg FACS), provide a classification for the six basic human emotions. Find effective methods for classification in machine learning and artificial neural networks. Disclose publicly available tutorial and test samples. Find acceleration options for new type of graphics processors.