Abstract
For the longest time, tennis analysis applications have always been extremely complex and expensive systems that only professional players and coaches could use and afford. This leaves the average amateur player in the dust; all they want to do is practice playing, but gaining access to this analysis, which includes costly multi-camera setups and advanced tennis expertise, would make this near impossible.
In comes Nice Shot!, an online platform which uses high-end Computer Vision and Machine Learning to help the everyday user easily gain access to this complex analysis by simply uploading a video of themselves attempting various tennis shots. The system will then provide the user with metrics, shot classifications as well as a 3D interfaceable replication of their performance. Furthermore, users can feel free to use the tool in a professional setting, meaning fans can retrieve quick analytics on their favorite player’s shots.
Development on this project has led to the delivery of a web application which would make 3D tennis analytics more accessible and inexpensive than ever. The system is specifically made to tolerate a wide range of abilities, so people of all ages, young or old, can use it. Moreover, the analysis pipeline we’ve developed provides a foundation for other tennis analytics as developers can use the API for their own purposes.
Demo Video
Development Team
Project Management
Below is a Gantt Chart displaying the progress from the 4th of October 2021 until the 30th of March 2022. Due to communication issues, the client requirements were only confirmed on Week 6, which delayed the start of the project.
