A Modular Approach to Facial Expression Recognition

We study the use of multi-layer perceptrons in applying artifical learning to the recognition of emotional expressions from frontal images of human faces. The perceptrons are trained using per-pixel luma data from the images’ mouth and eye areas, and map the inputs to one of 6 emotions. We compare 3 different methods for processing input information: 1) one network module for all inputs; 2) one network module for both eyes, and one for the mouth; 3) one network module for the mouth, one for the left eye, and one for the right eye. Our results show that involving multiple modules leads to better results, resulting in an overall performance of 84% images classified correctly.