| Autors: Raynova, R. N. Title: Weight based model for emotion recognition in a multimodal agent based architecture Keywords: emotion recognition, multimodal architecture, weight based m Abstract: Expressing and perceiving emotions is a vital part of communication between people and an important advance in improving the human-machine interface. Efficient and robust recognition is based on input data from different modalities thus imposing the usage of multimodal recognition architecture and multimodal recognition frameworks. The paper presents a multimodal agent-based architecture for emotion recognition and suggests a weight based model for the influence of the input multimodal data on the output results and also defines the impact of several factors such as gender, age, ethic group and personality. The suggested weight based model considers the emotion detection as a complex non-linear system and can be used in a multimodal agent based architecture for empirical evaluation of the weight, errors and the unknown parameters. References Issue
Full text of the publication |
Вид: пленарен доклад в международен форум, публикация в реферирано издание, индексирана в Scopus
