articuno0789/Emotion-Recognition

The objective of this document is to apply different intelligent classification methods to the problem of classifying emotion in images and videos of human faces. This well-defined problem is complicated by natural and unnatural variation in people's faces, which requires that the classification algorithm distinguish the small number of relevant features from the large set of input characteristics. 4 different types of classifiers are used, namely KNearest Neighbors, Support Vector Machine with linear kernel, Support Vector Machine with RBF kernel and Support Vector Machine with polynomial kernel. These classifiers are used to recognize two facial expressions: smiling and not smiling of humans in still images and video in real time. The metrics of the classifiers are compared to judge the best classifier for the recognition of emotions by means of facial expressions. Although the real-time system can reliably classify some of the emotions, more work is needed to build a robust, real-time system that works outside of laboratory conditions.