Towards new Realms of Human Computer Interaction Using Affective Computing and Persuasive Technology

提出的研究计划探讨了人机交互（HCI）领域的两个新思想：情感计算和说服技术。这些范式考虑了人机交互过程中情感的双向交流。因此，它们涉及赋予计算机情感智力和说服能力，这两种属性以前只属于人类。***情感计算方法旨在使计算机不仅能够理解和响应传统的输入，而且能够理解和响应用户的影响。（注意，affect指的是感觉或情绪的体验。）事实上，该领域的范围很广，涉及计算机对人类情感的识别、解释、响应和模拟。***我们建议发展情感识别方法，包括分析可以从视听传感器推断的信息的几种模式。与之前的工作相反，我们打算在分析中使用从视频信号中提取的生理信息。我们假设这种方法将提高使用视听传感器的情感识别的准确性。***此外，我们建议使用将开发的情感识别方法来评估HCI质量。因此，我们打算开发一个可用性检验框架，以支持通过HCI诱发的情绪的实证测量。这将允许交互设计师更容易地识别引起用户负面情绪的问题区域。说服技术涉及使用计算机来改变用户的思维和行为方式。他们的目标是通过人机交互产生的社会影响，诱使人们积极改变自己的态度。***我们建议在游戏应用中整合有说服力的概念。特别是，我们对游戏场景的动态调节感兴趣，以有效地呈现有说服力的内容，同时考虑到用户的情感状态和环境背景。用户的情感状态可以反映他/她对特定说服策略的反应程度。***这项研究的长期目标是在人类和计算机之间建立一种进化的关系，这种关系模仿了人类之间交流的一些社会特征。短期目标可总结如下：***开发多模态机制，在不依赖生物传感器的情况下，通过音视频捕获有效检测情感信息。***设计使用情感信息进行经验可用性测试的方法，并开发一个完整的框架来支持这种评估形式。***构思能够通过调节游戏场景而动态适应玩家情感状态和情境的说服性游戏

The proposed research program explores two novel ideas in the realm of Human Computer Interaction (HCI): Affective Computing and Persuasive Technology. These paradigms consider the bi-directional communication of emotions between humans and computers during interaction. Hence, they involve giving computers emotional intelligence and persuasive abilities, two attributes that were previously reserved for humans. ***Affective computing methods are aimed at allowing computers to appreciate and respond not only to traditional inputs, but also to users' affects. (Note that affect refers to the experience of feeling or emotion.) In fact, the field has a wide scope and is concerned with the recognition, interpretation, response and simulation of human affects by computers. ***We propose the development of affect recognition methods that involve the analysis of several modalities of information that can be inferred from audio visual sensors. As opposed to previous works, we intend to use physiological information extracted from a video signal in our analysis. We hypothesise that this approach will increase the accuracy of affect recognition using audio visual sensors.***Also, we propose the use of the affect recognition methods that will be developed in the evaluation of the HCI quality. Hence, we intend to develop a usability inspection framework that supports the empirical measurement of emotions evoked through HCI. This would allow interaction designers to more easily identify problematic areas that provoke users' negative emotions.***Persuasive technologies are concerned with the use of computers to change how users think and behave. Their goal is to entice humans to positively change their attitudes through social influence as a result of human-computer interaction. ***We propose to integrate persuasive concepts within gaming applications. In particular, we are interested in the dynamic modulation of gaming scenarios to efficiently present persuasive contents while considering the user's affective status and environmental context. The user's affective status can reflect her/his level of response to a particular persuasive strategy. ***The long-term goal of this research is to attain an evolved relationship between humans and computers that mimics some of the social characteristics reserved for inter-human exchanges. The short-term objectives can be summarized as follows:*** Develop multi-modal mechanisms to effectively detect affective information through audio-video capture without relying on bio-sensors. *** Devise methods for empirical usability testing using affective information and develop a complete framework to support such form of evaluation.*** Conceive persuasive games that can be dynamically adapted to gamers' affective state and context by modulating the gaming scenario.**