Selection-Based Mid-Air Text Entry on Large Displays

Abstract

Abstract. Most text entry methods require users to have physical devices within reach. In many contexts of use, such as around large displays where users need to move freely, device-dependent methods are ill suited. We explore how selection-based text entry methods may be adapted for use in mid-air. Initially, we analyze the design space for text entry in mid-air, focusing on singlecharacter input with one hand. We propose three text entry methods: H4 MidAir (an adaptation of a game controller-based method by MacKenzie et al.[21]), MultiTap (a mid-air variant of a mobile phone text entry method), and Projected QWERTY (a mid-air variant of the QWERTY keyboard). After six sessions, participants reached an average of 13.2 words per minute (WPM) with the most successful method, Projected QWERTY. Users rated this method highest on satisfaction and it resulted in the least physical movement.

Vulture

“…Text entry is an important activity and supporting it in mid-air would be beneficial for a number of scenarios such as work in sterile conditions (e.g., operating theatres), in augmented reality (e.g., with Google Glass), and when writing on public displays. It has been shown that people can write in mid-air with devices such as game controllers and dedicated gloves [21], but also using their hands [12,18]. However, text-entry rates for mid-air interaction are low, around 13 [18] to 18.9 WPM [27]; the latter rate was obtained with tactile feedback on errors and no character production on errors.…”

“…It has been shown that people can write in mid-air with devices such as game controllers and dedicated gloves [21], but also using their hands [12,18]. However, text-entry rates for mid-air interaction are low, around 13 [18] to 18.9 WPM [27]; the latter rate was obtained with tactile feedback on errors and no character production on errors. Furthermore, most techniques support only single-character text entry (e.g., [18,21,27]).…”

TiTAN

“…It is worth noting that, unlike recent unconstrained text entry evaluation [24], their system does not allow further inputs upon error occurrence, with the addition of tactile feedback. Using IR reflective markers tracked by multiple infrared cameras (Vicon/Optitrack), Markussen et al [11] were able to achieve 9.5 WPM on the first session and 13.2 WPM by the sixth session. Taking a more affordable approach, Ren et al [16] prototype a midair typing solution using a Microsoft Kinect and were able to achieve 6.11 WPM by the first day and 8.57 WPM on the fifth day.…”

“…Bi-manual Hunt-and-Peck Typing Extending on previous works on selection-based text entry [5,11,16,17,18], our technique supports bi-manual entry utilizing both hands while without requiring holding controller or marker. The user controls two on-screen pointers that hover on a virtual keyboard.…”

“…Midair interaction [1,5,7,12,14,17,18,20,22] is an emerging input modality for HCI, in particular for scenarios involving public displays [11,18], large surface [15], augmented and virtual reality [1] and sterile conditions (e.g., surgery). However, midair text entry has not received as much attention with most solutions requiring either the use of an external controller [1,6,14] or reflective markers [12].…”