Facial Emotion Recognition Using Fuzzy Systems
Austin Nicolai, Anthony Choi · 2015
This paper presents a fuzzy logic based emotion recognition system. The system is comprised of an image processing stage followed by an emotion recognition stage. In the image processing stage, the subject's face and facial features (eyes, mouth, etc.) are extracted. Next, the relevant identifying points are extracted from each facial feature. In the emotion recognition stage, the identifying points are used to fuzzify and determine the strength of different facial actions. These strengths are then used to determine the subject's displayed emotion. The Japanese Female Facial Expression database was used to evaluate the system's performance resulting in an overall successful detection rate of 78.8%.