Problem

For sports enthusiasts, analyzing player movements and performance is crucial in understanding the game and improving strategies. However, manually tracking player movements can be time-consuming and prone to errors.

To solve this problem, our team was tasked with implementing player movement tracking and tennis action recognition using Deep Neural Networks.

Click here

Solution

Our team utilized image processing techniques to extract relevant features from video footage of sports events. We then used TensorFlow and Mediapipe to train Deep Neural Networks to recognize specific actions such as serves, volleys, and forehands in tennis.

We also implemented Convolutional Neural Networks (CNN) for player movement tracking. This allowed us to accurately track the position of players on the field or court throughout the game.

Impact

Improved accuracy of player movement tracking by X%
Increased understanding of player performance and strategy through detailed analysis

Quote of how incredible of an impact this had.

Team

Team Member #1 - Data Scientist
- Led development of image processing techniques for feature extraction
- Implemented LSTM for action recognition in tennis footage
Team Member #2 - Machine Learning Engineer
- Developed CNN for player movement tracking
- Integrated TensorFlow and Mediapipe for training Deep Neural Networks

Tennis Action Recognition & Tracking with DNNs