Abstract

The Leap Motion Controller is a small USB device that tracks hand and finger movements using infrared LEDs, allowing users to input gesture commands into an application in place of a mouse or keyboard. This creates the potential for developing a general gesture recognition system in 3D that can be easily set up by laypersons using a simple, commercially available device. To investigate the effectiveness of the Leap Motion controller for hand gesture recognition, we collected data from over 100 participants and then used this data to train a 3D recognition model based on convolutional neural networks, which can recognize 2D projections of the 3D space. This achieved an accuracy rate of 92.4% on held out data. We also describe preliminary work on incorporating time series gesture data using hidden Markov models, with the goal of detecting arbitrary start and stop points for gestures when continuously recording data.

Conference Proceedings: , 2015 Conference on Modeling, Simulations and Visualization Methods