Research data supporting "HotGestures: Complementing Command Selection and Use with Delimiter-Free Gesture-Based Shortcuts in Virtual Reality"

1. The HotGestureData.zip contains a dataset used for training the recognizer in the HotGestures paper. The dataset consists of hand skeleton data of 10 gesture classes and the Null gesture class, collected from 8 participants. Each csv file is a gesture clip that records the trajectory of the 1) ABSOLUTE 3D position (x, y, z) and quaternion (x, y, z, w) of the wrist, and 2) positions and rotations of other joints RELATIVE to the wrist. Please read the markdown (.md) file for more details. 2. The HotGestureSource.zip contains the python source code of the recogniser. One can train this recognizer using the dataset above to implement the HotGestures.