Author(s)

Yachen Tian

Corresponding Author

Yachen Tian

Abstract

This study presents an intelligent assessment method based on multimodal data to enhance billiard skills. By collecting data from multiple sources, including video recordings, sensors, and physical parameters, the system comprehensively evaluates players' performance in key areas such as ball control, shot accuracy, positioning, and strategic thinking. A deep learning model with a Transformer architecture was employed to process and analyze the multimodal data, identifying strengths and weaknesses for each player. The system then generated personalized training programs tailored to the skill level of individual players. For beginners, the recommendations focused on improving basic posture and shot accuracy, while more experienced players received guidance on advanced tactical training. Experimental results demonstrated that the system significantly improved most participants' billiard skills in a short period. Additionally, the system held broader social value, especially in public welfare programs aimed at youth in underprivileged areas, helping them develop fine motor skills and strategic thinking through billiard education. This research introduces an innovative approach to evaluating and improving billiard skills by integrating intelligent technology and multimodal data analysis. Furthermore, it demonstrates the wide-ranging application of such technology in sports training, offering a novel method for accelerating skill development in both novice and advanced players, and highlighting the potential for expanding the system to other sports and educational contexts.

Keywords

Multimodal Data Fusion, Transformer Model, Billiard Skill Assessment, Sports Strategy Optimization, Tactical Decision-Making