TL;DR
Mobile touchscreens, like the iPhone, excel at navigation but struggle with object manipulation (cut, copy, paste, multi-select) without disrupting gestures. This paper introduces "Bezel Swipe," a novel technique that uses the device's physical bezel to initiate these operations, keeping them separate from standard zooming and panning. User studies show Bezel Swipe is a viable and well-received alternative to direct touch selection for these tasks.
What Problem Does This Paper Address?
Modern mobile devices, epitomized by the iPhone, have embraced touch interfaces with gestures like swiping and pinching for navigation (panning and zooming). While intuitive for browsing content, this design creates a conflict when users want to perform other common operations like selecting multiple items, cutting, copying, or pasting. These actions often require a mode change, which can be error-prone and disrupt the user's flow. The paper highlights that the iPhone, despite its popularity, lacks native cut and paste functionality, a direct consequence of this design challenge. The core problem is enabling these essential object manipulation tasks without interfering with the primary navigation gestures that users have come to expect.
Methods
The paper introduces Bezel Swipe, an interaction technique designed to overcome the limitations of direct touch selection on mobile devices. The core idea is to leverage the touch-insensitive bezel surrounding the display.
Here's how it works:
- Initiation: A Bezel Swipe gesture begins by touching the bezel.
- Entering Selection Mode: The finger is then swiped from the bezel into the display, which activates a selection mode.
- Object/Region Selection: The user then moves their finger over the desired object or region. Lifting the finger off the display concludes the selection mode.
- Action Differentiation: Different portions of the bezel can be marked to represent different actions (e.g., cut, copy, paste). In their prototype, a colored bar indicated these regions.
- Single vs. Multi-Object/Text Selection:
- Single Objects: A single swipe can select an object (like an image thumbnail).
- Text Regions: For text, a two-step process is used: the first swipe marks the beginning of the selection, and a second swipe marks the end, similar to Emacs' "mark and yank." This allows for selecting regions larger than the visible display without automatic scrolling.
- Verb-Noun Interaction: Bezel Swipe follows a "verb-noun" interaction pattern, where the gesture starts with the action (on the bezel) and ends with the object (on the display). This is contrasted with the more common "noun-verb" approach.
- Feedback: Continuous graphical feedback (a colored line from the activation bar to the fingertip) is provided to keep the user aware of the ongoing selection.
The technique is designed to be "conflict-free," meaning it doesn't interfere with standard zooming and panning gestures. It also aims to avoid the disadvantages of explicit mode changes by having the selection mode end when the gesture concludes.
Datasets
Not discussed in the paper. The paper describes an experimental evaluation using images of insects and faces presented in a grid on an iPhone display.
Key Insights and Ideas
- Leveraging the Bezel: The most significant insight is the clever use of the physical bezel, an often-underutilized area, to introduce new interaction capabilities without sacrificing screen real estate or disrupting existing gestures.
- Verb-Noun for Object Manipulation: Reversing the typical noun-verb interaction for object manipulation (starting with the action on the bezel) proves to be a natural fit for initiating these operations without conflicting with navigation.
- Implicit Mode Management: Bezel Swipe's selection mode is implicitly managed by the gesture itself, reducing the cognitive load and potential errors associated with explicit mode switching.
- Visual Feedback is Crucial: Providing continuous visual feedback during the swipe is essential for users to fine-tune their selections and feel in control.
- Viable Alternative to Direct Touch: The study results suggest that Bezel Swipe is not just a functional workaround but a genuinely preferred method for many users compared to more complex or less intuitive alternatives.
Key Results
The paper presents results from a user study comparing Bezel Swipe with three other methods: Direct (single tap selection), Single (mode-switching, single selection), and Multiple (mode-switching, multiple selections).
- Accuracy: All methods achieved similar high accuracy in selecting insect images (average 94.5%) and low error rates for selecting face images (average 0.43%).
- Task Completion Time:
- Direct was the fastest (101.34 sec).
- Multiple was the slowest (170.01 sec).
- Bezel Swipe (138.15 sec) and Single (140.66 sec) were in between.
- Errors:
- The Multiple method led to significantly more errors, particularly when users attempted to navigate while in selection mode.
- Accidental deselections were also more frequent with Multiple.
- Subjective Ratings (Figure 3):
- Bezel Swipe was rated as favorably as Direct on most measures, including "fun," "easy to use," and "requires a lot of interaction."
- Bezel Swipe was rated significantly more positively than Single and Multiple for these aspects.
- No significant differences were found for "tiredness" or "seeing UI elements."
- Direct was rated slightly higher for "requires high accuracy," but this didn't translate to significantly better performance.
Limitations and Future Work
- Limitations:
- Finger Occlusion: The authors acknowledge that Bezel Swipe, like other touch-based methods, is susceptible to finger occlusion, especially for text selection. They suggest combining it with techniques like "Shift" [12] to mitigate this.
- Text Selection Complexity: While Bezel Swipe can handle text selection, the paper notes that this is a more complex area and warrants further investigation.
- Future Work:
- Text Editing: Exploring Bezel Swipe for text editing tasks like highlighting, copying, and deleting.
- Scalability and Limitations: Further examining the technique's scalability and its limitations in various contexts.
- Device and Display Variations: Investigating its applicability on different types of devices (e.g., larger displays, multiple displays) and exploring extensions.
Summary updated