利用邊際感知設計使用者介面於多媒體互動

Abstract

這篇論文是在探討如何利用使用者的邊際感知（Peripheral Sense）來跟多媒體互動，所探討的範圍包括了圖形化介面（GUI, Graphical User Interface）、實體可觸性介面（TUI, Tangible User Interface）以及體感介面（KUI, Kinetic User Interface）。在圖形化介面中，我們藉著 SmartPlayer 以及 AmbientMemento 這兩個研究來探討如何藉由互動介面的設計，來利用使用者的邊際感知來瀏覽影像跟圖片。在實體可觸性介面中，iCon 以及它的延伸研究 MemoICON 則是設計如何利用使用者非常熟悉的辦公環境來達到利用邊際感知來利用桌面上的日常生活物品做為辦公電腦互動的延伸，來達到多工的可能性。在體感介面，PUB 以及其延伸研究 SonarWatch 則是探討當使用者在不用眼睛的互動情境下，怎麼樣利用自身體感來做操作，並且探討使用者在不用眼的情況下，利用前臂當輸入介面時，使用者認知的極限，跟互動的形態。藉由探討這三個不同領域的介面設計，我們可以結論出三個設計邊際感知與多媒體互動的原則，包括了：(1) 適合低精確度、低涉入度以及需要時常操作的任務；(2) 因為每個人的邊際感知的能力是不太一樣的，因此高度的客製化是需要的； (3) 使用者對於介面跟環境的熟悉程度，決定了邊際感知互動的可行程度。

This thesis explores how to utilize users’ peripheral senses, while interacting with the multimedia content in different interaction domains, which includes GUI, TUI and KUI. In GUI, SmartPlaer and AmbientMemento are designed to utilize users’ peripheral sense to design the new experience of video and photo browsing. In TUI, iCon and its extended system MemoICON are designed to let users use their peripheral senses to utilize their everyday objects as instant and alternative controllers in their familiar working environments. Users can use the everyday objects to perform multiple tasks, which have the characteristics, such as low precision, low engagement and high frequency. In KUI, PUB and its extended system SonarWatch is designed to let users use their forearm as the controllers while in eyes-free manner. The human factor is fully explored and the design guidelines are also extracted to know that users have their limitation to discriminate the number of discrete points on their forearms and moreover, different user has different tapped patterns, which also suggested highly customized calibration procedure is required. Summarizing the design guidelines in the three different interaction spaces, GUI, TUI and KUI, we can conclude that three important guidelines to design the interactions by utilizing users’ peripheral senses while consuming multimedia. (1) Low precision, and low engagement, high frequency tasks are suitable for the peripheral interactions. (2) Highly customization for peripheral interaction is required. (3) The more familiar interface can take more benefits of the peripheral senses.