Golf swing classification with multiple deep convolutional neural networks

Abstract

The use of smart sports equipment and body sensory systems supervising daily sports training is gradually emerging in professional and amateur sports; however, the problem of processing large amounts of data from sensors used in sport and discovering constructive knowledge is a novel topic and the focus of our research. In this article, we investigate golf swing data classification methods based on varieties of representative convolutional neural networks (deep convolutional neural networks) which are fed with swing data from embedded multi-sensors, to group the multi-channel golf swing data labeled by hybrid categories from different golf players and swing shapes. In particular, four convolutional neural classifiers are customized: ''GolfVanillaCNN'' with the convolutional layers, ''GolfVGG'' with the stacked convolutional layers, ''GolfInception'' with the multi-scale convolutional layers, and ''GolfResNet'' with the residual learning. Testing on the real-world swing dataset sampled from the system integrating two strain gage sensors, three-axis accelerometer, and three-axis gyroscope, we explore the accuracy and performance of our convolutional neural network-based classifiers from two perspectives: classification implementations and sensor combinations. Besides, we further evaluate the performance of these four classifiers in terms of classification accuracy, precision-recall curves, and F1 scores. These common classification indicators illustrate that our convolutional neural network-based classifiers can basically group the golf swing predefined by the combination of shapes and golf players correctly and outperform support vector machine method representing traditional classification methods.

A Random Forest-Based Accuracy Prediction Model for Augmented Biofeedback in a Precision Shooting Training System

“…Advances in science and technology today provide the means to improve and develop training techniques. In recent years, many sensors such as accelerometers [1,2], gyroscopes [3,4], and pressure sensors [5,6] have been widely used in various sports activities. For example, ultrasound sensors, which represent a reasonable compromise between cost and accuracy, have been used in archery to analyze the stability of the archer's hands [7].…”

Body sensor networks for monitoring performances in sports: A brief overview and some new thoughts

“…In Ref., [30] 19 different types of golf swing (straight, pull, push, slice, draw, hook, fade, etc.) are classified using accelerator, gyroscope and strain gage sensors (attached to golf club) based on various representative convolutional neural networks, i.e.…”

“…SVM classifier for step recognition [19] Grind tricks, Air tricks Preprocessing, threshold analysis for trick event detection, and classification of trick category using different types of classifiers [29] Canoe sprint phases Time series segmentation, feature extraction for describing motion phases, and SVM classification for motion phase [30] Different types of golf swing(straight, pull, push, Preprocessing by data augmentation, data shuffling, and data slice, draw, hook, fade, push-slice, pull-hook, etc.) standaridization followed by classification with customized deep CNNs [31] Landing momentum Preprocessing the data by a sensor calibration and alignment; Landing velocity calculation to measure the momentum of landing phase [32] Horse gaits (Walking, Rising trot) Sensor error calibration; Calibration of equestrian poster based on A-POS;…”

“…In Table 6, sampling rates used in sports performance monitoring are demonstrated, where it can be seen that the sampling frequency of 100 Hz is mostly used for the wearable sensors. [20,30] In Table 7, the problems associated with various studies together with some corresponding notes are presented. As we see, various issues like data transmission, sensors positions, sensors fusion, complexity, flexibility, etc.…”

“…Cloud assisted BSN approach can be exploited to solve real-time remote data transmission to provide an on-line motion monitoring system. [30] The computational complexity is not addressed. Representative Deep CNN-based advanced classifiers are used.…”