In this study, 3-year-old, 4-year-old, and 5-year-old children in three facilities (n=101) participated in the movement analysis during musical expression utilising a 3D motion capture system (MVN). Quantitative analysis, such as a three-way non-repeated analysis of variance of the captured data, yielded feature quantities that were implemented in machine learning algorithms with multiple classifiers. Results of classification regarding the developmental degree of musical expression in early childhood indicated higher accuracy with MLP-NN and SVM. Furthermore, using an eye tracking system (Tobii Glass 3) connected with MVN, simultaneous analysis of eye and body movements during musical expression improved the accuracy of MLP-NN to 74.42%. Measurements were conducted indoors at the facility under the same lighting conditions as the children's daily lives, so there was almost no effect of ambient light during eye tracking. Based on these findings, the MVN system, connected to Meta gloves, was employed to conduct detailed movement analysis focusing on hand movements during musical expression with 3-year-old, 4-year-old, and 5-year-old children in three facilities (n=86). Meta gloves, developed by MANUS Technology Group, are a small and lightweight system weighing 70 grams, which places almost no strain on the finger movements of the participating children. This system has a finger sensor that detects the absolute value of the fingertip position and the three-axis rotation angle of the finger relative to a model of the back of the hand at a frequency of 120 Hz. Because the Meta gloves are small and lightweight, the burden on fine motor control was suppressed. Quantitative analysis of this data revealed that the third metacarpal bone plays a dual role: supporting finger movements and activating the proximal and distal phalanges to facilitate the expression of imagined imagery. Admittedly, methodological limitations, such as lighting conditions and small, lightweight devices having some effect on the acquired data, are an area for further consideration. Utilising newly extracted feature quantities from this quantitative analysis is expected to improve the accuracy of classifying the developmental degree of musical expression. This, in turn, will provide objective criteria for evaluating developmental progress and necessary musical experiences, contributing to improvements in early childhood music education practices.
Author (s) Details
Mina Sano
Tokoha University, Shizuoka, Japan.
Please see the book here:- https://doi.org/10.9734/bpi/nhstc/v1/5557
