WUDA: Visualizing and Transforming Rotations in Real-Time with Quaternions and Smart Devices

Fast forward

Full Video

Keywords

Mathematical Foundations & Numerical Methods; Machine Learning, Statistics, Modelling, and Simulation Applications

Abstract

The rising popularity of inertial sensing via smart devices is evident, finding use in multidisciplinary applications, such as gesture recognition and image stabilization. Amid the popularity of analyzing raw motion signals and Euler angles, we highlight quaternion rotations as a robust alternative for studying device orientation. Quaternions are a powerful mathematical tool for representing and affecting three-dimensional rotations. Their abstract nature, however, can make it difficult for researchers to visualize the data they provide. Mobile sensing rarely invents practical tools to experiment with this mathematical method. Since visualization effectively communicates data findings, we develop an open-source, real-time app, Wuda, that allows users to observe the orientation of smart devices via inertial sensing. Wuda helps users reduce dimensional complexity and intuitively study quaternion transformations. We demonstrate the practical nature of Wuda in the context of fitness tracking.