Automatic Emotion Recognition Techniques

Facial emotions are the powerful, natural and immediate means of nonverbal communication for humans. There are six basic emotions which are anger, disgust, fear, happiness, sadness, and surprise. Automatic emotion recognition is interesting and challenging and has applications in areas like artificial intelligence, computer vision etc. Emotions can be recognized using different modalities like facial expressions, body movements, gestures or speech. To extract facial features from an image geometric feature-based and appearance based methods can be used. For the emotion recognition purpose various machine learning algorithms have been already applied e.g. SVM, decision trees, linear discriminant analysis, Bayesian networks, naive Bayes, neural networks. For the effective emotion recognition for real-life applications a combination of adaptive classifiers can be used. For speech, the prosody-pitch, intensity and duration- carry information related to emotions. This paper presents a review on various automatic emotion recognition techniques which uses one of the modalities like facial expressions, brain activities or speech.