Back-of-device tap recognition by leveraging magnetometer in commodity smartphones

Tap inputs on the back of smartphones provide a solution to the occlusion problem but come with challenges such as reduced accuracy and constrained functionality. In this work, we introduce MagTap, a novel magnetic-based input interface for smartphones, specifically leveraging magnets and ferromagnetic materials to recognize tap inputs made on the back of the smartphones. Through our evaluation with a MagTap prototype on an Apple iPhone 12, we demonstrate that MagTap can achieve high localization accuracy (e.g., an average accuracy of 95.3% for single taps and 88.2% for double taps), not losing usability and deployability.