Dynamic Arm Gesture Recognition Using Spherical Angle Features and Hidden Markov Models

Hyesuk Kim,Incheol Kim

doi:10.1155/2015/785349

Abstract

We introduce a vision-based arm gesture recognition (AGR) system using Kinect. The AGR system learns the discrete Hidden Markov Model (HMM), an effective probabilistic graph model for gesture recognition, from the dynamic pose of the arm joints provided by the Kinect API. Because Kinect’s viewpoint and the subject’s arm length can substantially affect the estimated 3D pose of each joint, it is difficult to recognize gestures reliably with these features. The proposed system performs the feature transformation that changes the 3D Cartesian coordinates of each joint into the 2D spherical angles of the corresponding arm part to obtain view-invariant and more discriminative features. We confirmed high recognition performance of the proposed AGR system through experiments with two different datasets.

Highlights

Gestures are a powerful human-to-human communication modality and the expressiveness of gestures allows for the altering of perceptions in human-computer interaction [1]
The arm gesture recognition (AGR) system learns the discrete Hidden Markov Model (HMM), an effective probabilistic graphical model for gesture recognition, from the dynamic pose of the arm joints provided by the Kinect API
Major approaches to vision-based gesture recognition include support vector machine (SVM), Dynamic Time Warping (DTW), and Hidden Markov Models (HMMs)

Summary

Introduction

Gestures are a powerful human-to-human communication modality and the expressiveness of gestures allows for the altering of perceptions in human-computer interaction [1]. Vision-based gesture recognition technology can be applied to multiple fields including human-robot interaction [2], computer game [3], sign language understanding for the hearing-impaired [4], and other fields [5,6,7]. With the release of lowcost 3D sensors like Kinect, the dynamic gesture recognition technology has gained increased attention. We introduce a Kinect-based arm gesture recognition (AGR) system design. The AGR system learns the discrete Hidden Markov Model (HMM), an effective probabilistic graphical model for gesture recognition, from the dynamic pose of the arm joints provided by the Kinect API. Because the variance of Kinect’s viewpoints and the different length of the subject arms can significantly affect the estimated 3D pose of each arm joint, it is difficult to recognize gestures reliably with these features. In order to overcome this problem and obtain view-invariant features, the AGR system performs the feature transformation that changes the

Related Works

Arm Gesture Recognition System

Experiments

Findings

Conclusions