A Gesture Controlled Music Playback System Using Convolutional Neural Network

摘要

In this paper, we propose an application of applying hand gesture recognition using Convolutional Neural Network(CNN) for music playback control. Our system begins with live image capturing from a USB camera; subsequently, followed by the initialization, calibration, and motion detection stages. To train the neural network, a set of segmented and tagged hand gesture images were used. According to our experimental results, our method has the advantages of low latency and high accuracy, which is capable of detecting a set of six gestures in real-time.