0 Interactive Technology: Teaching People with Autism to Recognize Facial Emotions José C. Miranda, Tiago Fernandes, A. Augusto Sousa and Verónica C. Orvalho Instituto Telecomunicações, Universidade do Porto, Instituto Politécnico da Guarda Portugal 1. Introduction In daily life, we interact with others by exchanging a huge quantity of information, including our current states of emotions, through facial expressions. Thus, faces are crucial for the recognition and understanding of emotions and for assisting communications and interactions between people. Individuals with autism tend to avoid looking at others’ human faces and find it hard to recognize facial expressions and emotions in themselves and in others (Baron Cohen, 1995). This incapacity to read emotions on the human face impairs their ability to communicate with other people (Baron-Cohen et al., 2007). Previous work has shown that children and adults with Autism Spectrum Disorders (ASD) can improve their emotion recognition skills with computer-based intervention. An increasing number of studies show that computer technologies used in teaching and in therapy are well accepted by individuals with ASD (Golan et al., 2007; Moore et al., 2000; Tanaka et al., 2010). However, most of the current computer-based solutions train patients with ASD using drawings or photographs (Bernard-Opitz et al., 2001; Bolte et al., 2002; Tseng & Do, 2010). When the autistic people try to apply the acquired knowledge to recognize emotions in real life, they are still unable to communicate properly (Golan et al., 2006). This paper gives an overview of existing methods that have been used for teaching emotion recognition to individuals with autism. We identify some technological limitations that difficult their interpersonal interactions. Lastly, we present our contribution: a different approach to teach autistic people to recognize emotions from facial expression. Our idea is based on real-time facial synthesis of 3D characters. We also suggest a different interaction model to involve the autistic patient more deeply in the process of learning emotions. Creating a solution to solve this problem requires a joint effort from many research fields, such as computer vision, computer graphics, human computer interaction and facial behaviour and emotions. 2. Emotions and facial expression Facial expression should not be confused with emotions. Human emotions are a result of many different factors, like feelings or convictions (Ekman, 1999). Human beings’ emotional states might be revealed through a number of different channels, such as emotional voice, pose, gestures, gaze direction and facial expressions. Facial expressions deal with facial 15 www.intechopen.com