Research And Applications Of Facial Expression Recognition Based On Locations Of Key Parts

M-Learning (Mobile Learning), a new E-Learning model, has becomemore and more popular, with the help of rapidly developed mobiletechnologies. It brings students tremendous convenience, but at meantime,poses new challenges. The lack of feedback has been a tough problem inE-Learning for teachers and students are separated in time and space. It iseven more difficult in M-Learning, especially detecting the learning stateswhich can be used to adapt the teaching strategy to improve the learningperformance. As one form of body language, facial expression is soeffective to reflect human’s inner state and emotional state that it plays asignificant role in human communication. Taking photographs of thestudent faces from mobile devices, recognizing their facial expressions andsending them back to the learning system or teachers is potentially helpfulto solve the problem mentioned above.Generally facial expression recognition involves three procedures: facedetection, expression features extraction and expression classification. Thisthesis introduces the research and application of facial expressionrecognition system in M-Learning. The main work includes:1. Locatingkey facial positions based on the improved Active Shape Model.2.Analyzing the relationship between features around key facial parts and theemotion states of students, extracting improved facial shape features andfacial appearance features, and classifying students’ emotion states withSupport Vector Machines.3. Conducting experiments differentiatingperson-relevant and person-irrelevant situations with the recognition rates 92%and78%respectively.4. Developing a facial recognition system inmobile learning situation based on client-server architecture.