핑거터치 제스처를 이용한 스마트 스피커 인터랙션

Background The smart speaker market is growing significantly every year, but its usability still does not meet the expectations of users. Voice interaction with existing smart speakers requires a pre-process consisting of a user's call and the smart speaker's response on the call, which has a limitation in that it is difficult to control errors immediately. Therefore, a new interaction is required that allows for more immediate error control while shortening the pre-call process of the existing smart speaker. Methods First, twelve unit operations essential for using the main functions of the smart speaker were derived, and various touch gestures posed by the user's one hand were devised. Second, a survey was conducted to decide a suitable gesture for each unit operation, and a smart speaker interaction based on the finger touch gestures was designed. Third, in order to evaluate the ease of learning and usability of the designed gesture interaction, usability evaluation was performed using the Wizard of Oz technique. Results As a result of the usability evaluation, the participants rated the use of finger touch gestures as useful, and almost all gestures were evaluated for fast learning and high usability. However, the one-tap gesture and the rubbing gesture on the top were found to have learning and usability problems, which needed improvement. Conclusions Although there is a limitation that the touch gesture can be used only when the smart speaker is in a position close to the user, such as at a bedside or on a desk, it can be operated very quickly without a pre-call process.