Abstract:
Facial expression is most expressive and understandable form of human communication to convey information. This study focuses on developing a model for facial expression recognition systems applying novel machine learning techniques. Speeded Up Robust Feature has been used for feature detection and description, Support Vector Machine used for classification purpose. The main top level procedures of this work are: Face Acquisition, Facial Expression Extraction (or feature extraction) and Expression recognition. Finally, this study comes up with hopeful facial image processing capability and accurate facial expression recognition based on the parameters. 92.79% on average accuracy has been achieved using JAFFE database considering different subject as test and 65.32 % has been achieved using YK database.