Predicting tap locations on touch screens in the field using accelerometer and gyroscope sensor readings
Published in Paper presented at the International Conference on Human-Computer Interaction, July 2020, 2020
Schmitt, E. & Voigt-Antons, J.-N.
We propose a lightweight method to infer tap locations on smartphones from inertial sensor data alone. Models trained on accelerometer and gyroscope signals generalize across users and devices and enable input reconstruction in real-world settings. Potential applications include implicit error detection, accessibility aids, and interaction analytics without screen capture.
Recommended citation: Schmitt, E. & Voigt-Antons, J.-N. (2020, July). Predicting tap locations on touch screens in the field using accelerometer and gyroscope sensor readings. Paper presented at the International Conference on Human-Computer Interaction. https://doi.org/10.1007/978-3-030-50309-3_43