A number of researches revealed the advantages of bezel-initiated gestures for mobile interaction. For example, Roth et al. (Roth & Turner, 2009) introduced Bezel Swipe for secondary actions like multi-target selection, copying and pasting. It was a simple single-finger swipe from the bezel towards the center of the touchscreen. They evaluated its usefulness in the context of image selection. Their participants found it to be a viable alternative to direct touch selection.
Bragdon et al. (Bragdon et al., 2011) prototyped Bezel Marks and Bezel Paths, which were single-finger gestures that start from the bezel, followed by mark-based and free-form path gestures, respectively. They found that bezel-initiated gestures were more resistant to situational impairments than soft-button-based interfaces. Serrano et al. (Serrano, Lecolinet, & Guiard, 2013) presented Bezel-Tap, which was a single-finger tap on the bezel followed by a single-finger swipe in the screen. It demonstrated the benefit of using sequential gesture combinations for designing new gestures. Bezel-initiated gestures have also been explored to address the limited thumb reach problem (Kim, Yu, & Lee, 2012; Li, Fu, & Zhu, 2016; Yu, Huang, Hsu, & Hung, 2013). Those works used a bezel-initiated gesture to activate a screen cursor for object selection, or to transform the screen space to bring faraway object closer to the thumb. They all used simple single-finger bezel-initiated swipe, without any suggestion on how to explore the design space and design bezel-initiated gestures. It is not obvious whether initiating from the bezel would be a limiting design constraint to interaction designers or end users. Instead, our work is not about the design of a specific bezel-initiated gesture for a specific task but the exploration of the multi-dimensional gesture design space.