In the field of sports training, practice plays a crucial role in improving performance. However, the ability to mimic the techniques of professional athletes can take a player’s skills to the next level. Personalized sports coaching assistants powered by artificial intelligence (AI) can make this a reality by utilizing published datasets. These systems use cameras and sensors placed on the athlete’s body to track movements such as joint patterns, muscle activation levels, and gaze movements. With this data, personalized feedback on player technique and improvement recommendations can be provided, making these systems versatile for athletes at all levels.
In a study published in Scientific Data on April 5, 2024, researchers led by Associate Professor SeungJun Kim from the Gwangju Institute of Science and Technology (GIST) and researchers from Massachusetts Institute of Technology (MIT) have developed a MultiSenseBadminton dataset for AI-driven badminton training. There is a lack of comprehensive badminton action datasets for analysis and training feedback, making this dataset a valuable contribution to the field. The study was supported by the GIST-MIT project and took inspiration from MIT’s ActionSense project, focusing on capturing movements and physiological responses of badminton players to enhance the quality of forehand clear and backhand drive strokes. The researchers collected 23 hours of swing motion data from 25 players with varying levels of training experience, using sensors to track joint movements, muscle signals, foot pressure, and shuttlecock positions.
The MultiSenseBadminton dataset aims to build AI-based education and training systems for racket sports players. By analyzing disparities in motion and sensor data among different levels of players, AI-generated action trajectories can be created to provide personalized motion guides for each player level. This dataset can enhance training through haptic vibration or electrical muscle stimulation, promoting better motion and refining swing techniques. Additionally, the player tracking data in the dataset could be used for virtual reality games or training simulations, making sports training more accessible and affordable.
In the long run, the researchers believe that the MultiSenseBadminton dataset could make sports training more accessible and affordable to a broader audience, promote overall well-being, and foster a healthier population. By using this dataset, individuals can improve their sports skills through personalized feedback and training methods, potentially transforming how people exercise and engage in physical activities. The data collected in the study can be used to develop AI models that evaluate stroke quality and provide feedback to players, enhancing their overall performance.
Overall, the development of the MultiSenseBadminton dataset represents a significant step towards leveraging AI technology to improve sports training and performance. By utilizing wearable sensors and cameras, researchers were able to capture detailed movement data from badminton players, allowing for the creation of personalized training guides and feedback. This dataset has the potential to revolutionize the way athletes train and improve their skills, making sports training more accessible and affordable for individuals of all levels. The researchers hope that this dataset will contribute to promoting physical well-being and healthier lifestyles among the population.
