Affective-aware tutoring platform for interactive digital television

Interactive Digital TeleVision (IDTV) is emerging as a potentially important medium for learning at home. This paper presents a novel affective-aware tutoring platform for IDTV which makes use of automatic facial emotion recognition to improve the tutor-student relationship. The system goes further than simply broadcasting an interactive educational application by allowing the personalization of the course content. The tutor can easily access academic information relating to the students and also emotional information captured from learners' facial expressions. In this way, depending on their academic and affective progress, the tutor can send personal messages or extra educational contents for improving students' learning. In order to include these features it was necessary to address some important technical challenges derived from IDTV hardware and software restrictions. The system has been successfully tested with real students and tutors in a non-laboratory environment. Our system tries to advance in the challenge of providing to distance learning systems with the perceptual abilities of human teachers with the final aim of improving students learning experience and outcome. Nevertheless, there is still relatively little understanding of the impact of affect on students' behaviour and learning and of the dynamics of affect during learning with software. Systems like ours would make it possible to attack these relevant open questions.