A Neural-Network-Based Approach to Adaptive Human Computer Interaction

Abstract

A neural-network-based approach is proposed in this paper providing multimedia systems with the ability to adapt their performance to the specific needs and characteristics of their users. An emotion recognition system based on facial cues is used to illustrate the performance of this approach. Although some archetypal emotions are considered to be universal, there is no doubt that the way they are expressed diversifies from one individual to the other. The initial generic system is adapted in order to meet the specific user’s traits.