Object Oriented Design for Multiple Modalities in Affective Interaction
The purpose of this chapter is to investigate how an object oriented (OO) architecture can be adapted to cope with multimodal emotion recognition applications with mobile interfaces. A large obstacle in this direction is the fact that mobile phones differ
- PDF / 529,301 Bytes
- 13 Pages / 439.37 x 666.142 pts Page_size
- 33 Downloads / 180 Views
Object Oriented Design for Multiple Modalities in Affective Interaction
Abstract The purpose of this chapter is to investigate how an object oriented (OO) architecture can be adapted to cope with multimodal emotion recognition applications with mobile interfaces. A large obstacle in this direction is the fact that mobile phones differ from desktop computers since mobile phones are not capable of performing the demanding processing required as in emotion recognition. To surpass this fact, in our approach, mobile phones are required to transmit all data collected to a server which is responsible for performing, among other, emotion recognition. The object oriented architecture that we have created, combines evidence from multiple modalities of interaction, namely the mobile device’s keyboard and the mobile device’s microphone, as well as data from user stereotypes. All collected information is classified into well-structured objects which have their own properties and methods. The resulting emotion detection platform is capable of processing and re-transmitting information from different mobile sources of multimodal data during human–computer interaction. The interface that has been used as a test bed for the affective mobile interaction is that of an educational m-learning application.
8.1 Overview of the Emotion Recognition System’s Architecture The authors of Neerincx and Streefkerk (2003) describe a study where emotion, trust and task performance are investigated as important elements of user interaction with mobile services. The participants of this study performed interaction tasks with mobile services, using small handheld devices and laptops. This study concludes with the presentation of the relations between trust, performance, devices and emotions of the users. In Gee et al. (2005) mobile telephones were used to collect data in order to find the relationship between gambling and mood state from gamblers. The results of this study revealed that subjective anxiety/arousal levels
E. Alepis and M. Virvou, Object-Oriented User Interfaces for Personalized Mobile Learning, Intelligent Systems Reference Library 64, DOI: 10.1007/978-3-642-53851-3_8, Springer-Verlag Berlin Heidelberg 2014
87
88
8 Object Oriented Design for Multiple Modalities
were significantly higher during and after gambling than during the urge to gamble. The authors of this study also state that collecting data through the use of mobile telephones appeared to be a valuable development in their research. Isomursu et al. (2007) collect affective interaction data that emerge from mobile applications using several emotion collection methods. As a next step, these methods are evaluated and the authors discuss about the experiences they have gained using these methods, and also provide a comparison framework to summarize their results. These studies provided strong evidence that emotion recognition in mobile interaction is very important, though it is difficult to be achieved. In this section, we describe the general architecture of the out e
Data Loading...
