Intelligent Table Tennis Coach

Table tennis has been steadily growing in popularity in the United States, but the high cost of hiring a personal coach has prevented many players from receiving the training they need to improve their skills. To address this issue, we have developed the Intelligent Table Tennis Coach, an innovative system that provides affordable and effective training for players of all levels. Our Intelligent Table Tennis Coach utilizes cutting-edge technology to teach players the fundamental elements of the game, including consistency, framework, timing, and acceleration. By using poses extracted by mediapipe [2], color recognition based on OpenCV [3] to detect the ball's direction, and the ADXL345 sensor to measure acceleration, our system can evaluate a player's performance from multiple angles and provide personalized feedback to help them improve. As far as we know, Intelligent Table Tennis Coach is the first table tennis training system that incorporates pose estimation, ball tracking, and acceleration evaluation. By providing a comprehensive evaluation of a player's performance, our system enables players to identify their strengths and weaknesses and develop their skills in a fun and engaging way. Unlike traditional coaching methods that can be expensive and time-consuming, our Intelligent Table Tennis Coach offers an affordable and convenient alternative. Whether you're a beginner or an experienced player, our system can help you improve your game and take your skills to the next level. Give it a try today and experience the benefits for yourself!

Motivation

Table tennis is a popular sport in the United States, but the high cost of hiring a personal coach has created a barrier for many players looking to improve their skills. Through our research, we discovered a significant demand for affordable table tennis coaching. We were inspired to develop a system that would be both accessible and effective in helping players improve their game. Drawing on our expertise in IoT and our shared passion for table tennis, we developed the Intelligent Table Tennis Coach. Our goal was to create a system that would provide amateur players with access to personalized coaching and feedback, without the high costs associated with traditional coaching methods. By combining our knowledge of IoT technology with our love for table tennis, we were able to design a system that incorporates pose estimation, ball tracking, and acceleration evaluation to provide comprehensive feedback on a player's performance. Our system is not only easy to implement, but also affordable, making it accessible to a wide range of players. Our motivation for developing the Intelligent Table Tennis Coach was to make table tennis coaching accessible to all players, regardless of their financial resources. We believe that our system has the potential to revolutionize table tennis training and enable players to reach their full potential. Give it a try and see for yourself how the Intelligent Table Tennis Coach can help you improve your game!

System

Architecture

Technical Components

Mediapipe, OpenCV, Pose estimation, Color recognition, Flask server

Prototype

Results

References

[1] Zhang, Z. (2022). Action Recognition and Application of Table Tennis Training Based on IOT Perception. Security and Communication Networks, 2022.

[2] Lugaresi, C., Tang, J., Nash, H., McClanahan, C., Uboweja, E., Hays, M., ... & Grundmann, M. (2019, June). Mediapipe: A framework for perceiving and processing reality. In Third Workshop on Computer Vision for AR/VR at IEEE Computer Vision and Pattern Recognition (CVPR) (Vol. 2019).

[3] Bradski, G., & Kaehler, A. (2000). OpenCV. Dr. Dobb’s journal of software tools, 3, 120.

Contact

Zihao Huang: zihao.h@columbia.edu
Yukai Song: ys3493@columbia.edu
Kun Wang: kw2962@columbia.edu

Columbia University Department of Electrical Engineering
Instructor: Professsor Xiaofan (Fred) Jiang