A Comparative Evaluation Of Pen And Finger Gestures In Mobile Environments

In recent years, smart devices have been increasing in popularity and ubiquity. There is apressing need to find more effective user interfaces to help people to interact with smartdevices. As next generation user interface in human computer interaction, natural userinterface is very important for us to exploit and develop.One way for natural user interface is gesture interaction. When we interact with a touchbased smart device, drawing gestures is more convenient than clicking buttons or selectingitems from a long menu. In other words, gesture interaction is closer to our natural behavior.Pen and finger are the most familiar and popular input modalities in our daily life. Usinga pen or finger to draw gestures on the touch screen is a reasonable way in gesture interaction.On one hand, from childhood to old, we have a long time of using experience with a pen. Onthe other hand, the finger is a part of our body, it is the most natural way to touch smart devicewith a finger. Because of the strong points of pen and finger, they had already become themost important input modalities on the touch screen. In addition, finger interaction becameincreasingly popular in these years because of the development of capacitive touch screen.To advance pen and finger gesture interaction, a lot of studies have been conducted onpen and finger gesture. Consider about the development of smartphone and smart watch, notonly we can use them in static environments, they also can be used in mobile environments.Little study has been done in the comparative evaluation of pen and finger gesture in mobileenvironments. Thus, we conducted two experiments to examine the differences andsimilarities between pen and finger gesture while walking.Firstly, we summarized the study of pen and finger gesture in static environments. Thenwe did our gesture comparative experiment on a treadmill. Using three different inputmodalities (pen, index fnger and thumb), we investigated user performance while drawing aset of gestures with varying degrees of complexity and target sizes in terms of nine features.Some significant results have got in the experiment, but comparing it to the result in staticenvironments, some differences cannot be well understood. At last, to check the results andfurther understand about real walking condition, we designed another experiment in a real walking environment. The participants drew the gestures on a smartphone while a pacesetterwalking side by side. In this way, the walking speed can be controlled in real walkingenvironment.The result revealed that from the static environment to real walking until walking on thetreadmill is a continuously varying process. The performance of all input modalities wasdegraded in mobile situations, but the performance of pen decreased faster than the fingers.Thumb always performed the worst because of its biological limitations. All in all, based onthe results, we provide design implications for mobile touch interfaces.