An Approach for Improving Automatic Mouth Emotion Recognition

摘要: The study proposes and tests a technique for automated emotion recognition through mouth detection via Convolutional Neural Networks (CNN), meant to be applied supporting people with health disorders communication skills issues (e.g. muscle wasting, stroke, autism, or, more simply, pain) in order recognize emotions generate real-time feedback, or data feeding systems. software system starts the computation identifying if face is present on acquired image, then it looks location extracts corresponding features. Both tasks are carried out using Haar Feature-based Classifiers, which guarantee fast execution promising performance. If our previous works focused visual micro-expressions personalized training single user, this strategy aims train also generalized faces sets.