NailRing: An Intelligent Ring for Recognizing Micro-gestures in Mixed Reality
Tianyu Li, Yue Liu, Shining Ma, Mingwei Hu, Tong Liu, Weitao Song
Abstract
Gesture interaction is currently a main interaction technology in the field of mixed reality. However, long-term and large-scale gesture in mid-air will lead to muscle fatigue and privacy problems, which cannot meet the comfort requirements of continuous interaction and inevitably hinder the development of mixed reality systems. To solve this problem, we propose NailRing, an intelligent ring to recognize fingertip micro-gestures using a micro-close-focus camera on a fingertip bracket. Such fingertip physiological characteristics as the changes in fingertip color distribution and muscle shape changes caused by fingertip pressure have been studied. According to the recognition principle, ten types of micro-gestures have been designed and used for contact interaction and one-hand interaction respectively. The accuracy of gesture recognition (cross-session $ F_{Macro}=98.3\%$; cross-person $ F_{Macro}=86.4\%$) in user studies verifies the performances of NailRing under different interaction conditions. Finally, the capability of NailRing in a series of potential application scenarios has also been discussed and analyzed.