GIST and MIT researchers have created the MultiSenseBadminton dataset to boost badminton training using AI.

The dataset encompasses 23 hours of swing motion data from 25 players of different skill levels.

Data collection involved sophisticated sensors like IMU, EMG, insole sensors, and video capture.

A machine learning model was employed to label and validate the data for AI training effectiveness.

The aim is to utilize the dataset to develop personalized training systems for racket sports.

Advanced training methods may include haptic vibration or electrical muscle stimulation.

The initiative could lead to more accessible and affordable sports training, promoting health and fitness.