RestGest: Resting Elbows for Sustainable Mid-Air Gesture Interaction

  • Lee, Donghun; 
  • Kim, Minhyeok; 
  • Lee, Euije; 
  • Lee, Junhyub; 
  • Oh, Sangeun; 
  • ... Kim, Hyosu
Citations

SCOPUS

0

초록

Natural interaction in daily life should let users control devices without extra effort or interruption. Mid-air hand gestures are attractive for this purpose, but they are often tiring, require dedicated or privacy-sensitive sensors, or support only a limited set of gestures. This work aims to make mid air gesture input less tiring and easier to deploy in everyday situations with a nearby surface available for elbow support. We present RestGest, which asks users to rest their elbows on nearby surfaces such as desks, tables, or armrests while gesturing. This simple posture reduces arm fatigue and allows gestures such as pinching and clapping to generate vibrations that travel through the arms and surfaces. An eight-participant feasibility study found that elbow support reduced perceived arm fatigue during mid-air interaction. RestGest captures these vibrations using only the built-in inertial sensors of a smartphone on the surface, a smartwatch on the wrist, or both. We design a practical gesture vocabulary and a recognition pipeline for everyday use. Across 15 participants, the default dual-device configuration achieved a 96.4% end-to-end command success rate. These results demonstrate the feasibility of ergonomic mid air gesture interaction using commodity devices.

키워드

Gesture recognition; Human computer interaction; Mobile computing
제목
RestGest: Resting Elbows for Sustainable Mid-Air Gesture Interaction
저자
Lee, Donghun; Kim, Minhyeok; Lee, Euije; Lee, Junhyub; Oh, Sangeun; Kim, Hyosu
DOI
10.1109/TMC.2026.3737144
발행일
2026
유형
Article
저널명
IEEE Transactions on Mobile Computing