Assisting Tennis Players
School STEM Fair - Gold Medal - 2022
About the Course
About this project
There are almost 500 times more tennis players than tennis coaches-- leaving many aspiring tennis players without a coach. Training with a coach can also be expensive, self-training can be flawed and with the shift to online-based activities due to the pandemic, tennis players need a solution that can still help enhance their serve and form while manouvering current restraints.
Advaith wanted to create an AI that could analyze a serve done by a tennis player through smartphone video footage, and then provide diagnostic feedback so that human intervention is not necessary. Advaith used a pre-existing CNN/RNN model to help draw joints and bones onto a body and trained the main algorithm. Through a website and app, users can upload video footage and the algorithm will process it and offer feedback.
This is a great low-cost and accessible solution for aspiring players to refine their skills as well as train by themselves without the need for a coach.