ClickSense: A Low-Cost Tangible Active User Input Method Using Passive Acoustic Sensing for Mobile Virtual Reality

Mobile virtual reality (MVR) employs smartphone-integrated head-mounted displays to provide low-cost immersive experiences widely available to the public. Compared to higher-fidelity devices, MVR offers poor interaction support notably lacking a mechanism to produce click events necessary to indicate selection. We propose ClickSense: a no-cost active input method based on acoustic sensing. ClickSense uses the smartphone's built-in microphone to passively listen to environmental audio to identify the unique click sound emitted from a handheld clicker. Keeping with the spirit of the low-cost nature of MVR, the clicker itself is made of recycled household objects for no cost. We present a sound recognition system that can easily run on lower-powered mobile devices. Unlike vision-based gestures, our approach is not dependent on the smartphone's camera fidelity. Our evaluation of the system reveals that it offers a 97.33% recognition rate when used in isolation. When used to indicate selection in a standard selection task, it offered a 95.05% recognition.