Autism is a communication disorder that mandates early and continuous educational interventions on various levels like the everyday social, communication and reasoning skills. Computer-aided education has recently been considered as a likely intervention method for such cases, and therefore different systems have been proposed and developed worldwide. In more recent years, affective computing applications for the aforementioned interventions have also been proposed to shed light on this problem. In this paper, we examine the technological and educational needs of affective interventions for autistic persons. Enabling affective technologies are visited and a number of possible exploitation scenarios are illustrated. Emphasis is placed in covering the continuous and long term needs of autistic persons by unobtrusive and ubiquitous technologies with the engagement of an affective speaking avatar. A personalised prototype system facilitating these scenarios is described. In addition the feedback from educators for autistic persons is provided for the system in terms of its usefulness, efficiency and the envisaged reaction of the autistic persons, collected by means of an anonymous questionnaire. Results illustrate the clear potential of this effort in facilitating a very promising autism intervention.

Abstract

2 SummaryBackground: Affective computing (AC) is concerned with emotional interactions performed with and through computers. It is defined as "computing that relates to, arises from, or deliberately influences emotions". AC enables investigation and understanding of the relation between human emotions and health as well as application of assistive and useful technologies in the medical domain. Objectives: 1) To review the general state of the art in AC and its applications in medicine, and 2) to establish synergies between the research communities of AC and medical informatics. Methods: Aspects related to the human affective state as a determinant of the human health are discussed, coupled with an illustration of significant AC research and related literature output. Moreover, affective communication channels are described and their range of application fields is explored through illustrative examples. Results: The presented conferences, European research projects and research publications illustrate the recent increase of interest in the AC area by the medical community. Tele-home healthcare, AmI, ubiquitous monitoring, e-learning and virtual communities with emotionally expressive characters for elderly or impaired people are few areas where the potential of AC has been realized and applications have emerged. Conclusions: A number of gaps can potentially be overcome through the synergy of AC and medical informatics. The application of AC technologies parallels the advancement of the existing state of the art and the introduction of new methods. The amount of work and projects reviewed in this paper witness an ambitious and optimistic synergetic future of the Affective Medicine field.

No abstract

This paper introduces initial attempts towards an emotion specification method for physiological signals. The core elements of a complete emotion record are identified and their structure is represented. The XML annotation is proposed as the most appropriate since it provides platform independency and easier intercommunication. The emotion record is described in both quantitative statistical and qualitative analytical manners. The former is a standard way of representing signals by specifying their true numerical values. The second, though is additional description that is characteristic for each signal and describes the main features of the signal in a more free and informal annotation.