Innovative toolkit for objective evaluation of traditional Thai dancing by intern teachers using Kinect, MediaPipe, and microcontroller technology

Abstract

To assess traditional Thai dance proficiency, evaluators often rely on subjective observations, which can be challenging for inexperienced intern teachers. This study aimed to create an innovative toolkit for intern teachers to more accurately evaluate their students’ traditional Thai dance proficiency. The objective was to develop and assess the toolkit’s precision, accuracy, and user satisfaction. Ten intern traditional Thai dancing teachers were selected as the sample group for this work. The innovated toolkit is comprised of Kinect v2, add-on camera with MediaPipe API, and a microcontroller connected to flex sensors sending signals to computing unit to automatically compute the scores. During the field usage, evaluation forms were also given to three traditional Thai dancing experts to evaluate the sample group as the innovated toolkit was used to evaluate the same sample groups simultaneously. The results from both evaluation methods were compared and the toolkit demonstrated 9.61% deviation from experts’ opinions giving the toolkit 90.39% accuracy. Through four repeated evaluations with one of the experts, the score fluctuated with standard deviation of 2.58 out of 100 giving the tool’s precision to be ±1.49%. This provides cost-effective and accurate evaluation platform catered towards basic traditional Thai dancing lessons.

Keywords

• Thai Dance
• Art Digitalization
• Kinect
• MediaPipe
• Microcontroller
• Collaborative Design