利用漸進式視覺引導以完成連續型視線輸入

Abstract

連續型視線輸入解決了以視線輸入時眼球顫動及誤觸操作的問題。過去的研究多將焦點集中在掃視式的視線輸入，透過將筆畫操作拆解為階段直線運動的方式來達到增進操作效率的目的，但為這種機制設計的視覺導引系統難以直接套用在任意的筆畫手勢上。因此我們結合連續型視線追蹤與動態視覺導引的概念，提出一個專為連續型視線輸入設計的漸進式動態視覺導引系統 GazeBeacon，能在操作過程中於視線輸入端點周圍持續地提供即時的視覺回饋與前饋。在前導實驗中我們發現兩個問題。我們利用套用平滑濾波器及輸入點重新取樣的兩個方式減緩眼球顫動隊對長度計算帶來的影響，並在路徑導引前方增加一視線集中焦點以解決使用者錯誤判讀引導的問題，設計一個基於不同筆畫組成類型探討的實驗來驗證這個做法。最後我們將 GazeBeacon 與傳統快捷查詢表做操作時間、操作辨識率與操作正確率的比較，發現雖然前者讓使用者花費更多時間，但也顯著地降低了在連續型視線輸入介面上的操作失誤率。

Gaze-gesture interaction solved the jittering and the Midas Touch problem of gaze-controlled interfaces. Previous works mainly focused on saccadic gaze gestures, increasing efficiencies by dividing gestures into segments. However, those guidance techniques could not directly be applied to any graffiti-like gestures. By combining the concept of smooth pursuit and dynamic guide, we propose GazeBeacon, a gradual visual guiding system designed for gaze-gesture interaction. It continuously provides real-time feedback and feedforward graphical cues under gaze points during the progress of the interaction. Two issues were found in our pilot study: miscalculation and misestimation. We mitigated the miscalculation problem by adding smoothing filters on gaze points and resampling the path before length calculation, while the misestimation problem was solved by adding a focus point at the end of the guidance path. These methods were verified by a user study based on different gesture primitives. Finally, we evaluated the completion time, the recognition rate and the selection accuracy of GazeBeacon, compared with the traditional crib-sheet guide. The result shows that although GazeBeacon makes users spent more time on execution, it significantly improves the accuracy of menu selections on gaze-gesture interfaces.