We present FinDroidHR, a novel gesture input technique for off-the-shelf smartwatches. Our technique is designed to detect 10 hand gestures on the hand wearing a smartwatch. The technique is enabled by analysing features of the Photoplethysmography (PPG) signal that optical heart-rate sensors capture. In a study with 20 participants, we show that FinDroidHR achieves 90.55% accuracy and 90.73% recall. Our work is the first study to explore the feasibility of using optical sensors on the off-the-shelf wearable devices to recognise gestures. Without requiring bespoke hardware, FinDroidHR can be readily used on existing smartwatches.
|Number of pages||42|
|Journal||ACM Proceedings on Interactive, Mobile, Wearable and Ubiquitous Technologies|
|Publication status||Published - Mar 2018|
- Wearable Computing
- Gesture Interaction
- Machine Learning
- Mobile Sensors