Touchless Gesture Recognition for Smart Mirrors

Internet of Things devices are used almost everywhere. “Smart mirror” devices are also gaining popularity. However, in most cases, operating the device requires contact with the surface of the mirror, which leads to surface contamination and the risk of spreading infections in public places. In order to solve this problem, the study proposes an algorithm for visual recognition of touchless gestures by a camera, based on machine learning. Experiments were conducted with the participation of subjects, and showed that the proposed algorithm demonstrates an accuracy of gesture recognition by a “smart mirror” of more than 90%. The paper also provides recommendations for ensuring information security of user interaction with the “smart mirror”.