PersonalTouch: Improving Touchscreen Usability by Personalizing Accessibility Settings based on Individual User’s Touchscreen Interaction

A user perform six different kinds of gestures and PersonalTouch analyzes user's input to recommend personalized, optimal accessibility settings including 2.00 seconds for 'Hold Duration' and 2.00 seconds for 'Tap Assistance with Initial Touch Location'.

Modern touchscreen devices have recently introduced customizable touchscreen settings to improve accessibility for users with motor impairments. For example, iOS 10 introduced the following three Touch Accommodation settings: 1) Hold Duration, 2) Ignore Repeat, 3) Tap Assistance Gesture Delay. These three independent settings lead to numerous possible configurations, making it impractical to manually determine the optimal setting. We present PersonalTouch, which collects and analyzes touchscreen gestures performed by individual users, and recommends personalized, optimal touchscreen accessibility settings.

We describe our suggestion pipeline as follow:

Three steps for PersonalTouch to provide recommendations. First, collect user's touchscreen gestures data. Secondly, simulate different combinations of settings on the input touchtracks. And finally, choose the optimal settings via cross-validations and settings responsiveness.

PersonalTouch goes through three steps to provide personalized settings suggestion, which includes a) Touch data collection. b) Gesture recognizer simulation and c) Settings recommendation.


Bar chart of the overall sucess rates for Default (55.7%) and PersonalTouch (75.9%) recognition models.

Our 12-motor-impaired-participant study shows that PersonalTouch significantly improved overall success rate by 20.2% over the default gesture recognizer.

For further information of our system and study, please refer to the paper.

Publication / Media

Competition achievement

This work won the overall 1st place for the 2019 undergraduate research competition held by CS department at National Taiwan University.