Embodied Interaction Design for Articulation Elicitation and Feedback

点击翻译按钮获取中文摘要

An increasing number of computer science graduates, as part of their employment, will become members of collaborative teams that are tasked with the development of interactive systems. The increasing accessibility of tools for digital media and software creation means that interactive systems are being developed for a variety of new domains and applications. The process of designing an interactive system that effectively accomplishes its objectives (such as to provide a certain type of user experience or to support learning) requires a multi-disciplinary repertoire of skills and capabilities. Increasingly, these systems go beyond mouse-and-keyboard interaction and use gestural or whole-body interaction (using sensor-based input). The development of such systems requires specialized knowledge about embodied interaction. In addition, such systems are most often developed iteratively, with a strong reliance on formative, empirical user evaluation at intermediate phases, and thus require specialized knowledge about evaluation methodologies. This program of research will investigate how the usability of articulation feedback systems is impacted by (1) the use of metaphorical, as opposed to realistic, visual representations of the vocal tract and (2) graphical perspective. Graphical perspective refers to the point of view with which a visualization is shown (such as from the embodied point of view that an in-world agent has or from a disembodied and more encompassing point of view). We will examine several aspects of usability, including the task outcome (effectiveness), the effort required to accomplish task outcomes (efficiency), and qualities of user experience, such as satisfaction and enjoyment. We will focus on users who are older adults. In undertaking this work, we will develop multiple new variants of interactive systems that will use electromagnetic articulography in order to track in real-time the user's tongue position during the production of speech sounds. This work will emphasize the importance of iterative design and on-going empirical testing and evaluation.