Sports SensAI
Sports SensAI is a Python Flask web application for analyzing basketball shooting techniques. Through AI and Machine Learning, this app is a platform that aims to enhance shooting skills and accuracy. My current motivation stems from my desire to leverage technology in basketball to benefit underprivileged children. Sports SensAI not only assists coaches in refining techniques but also provides opportunities for children to learn and advance in basketball. By collaborating with government schools and academies in Delhi-NCR, I aim to revolutionize the way mechanical skills are taught and promote inclusivity in sports education.
Impact
15 Coaches
15 basketball coaches at academies and government schools across Delhi-NCR have recognised and tested out my app as a useful, smart analysis tool. IB3L also lauded my work as a great initiative.
150+ Players
Across academies and schools, more than 150 players have been able to analyze their own shooting action. This has led to better feedback from coaches and an improvement in their techniques.
News Feature
Sports SensAI has been featured in an article and Instagram interview by Ekalavyas Basketball. Additionally, I had an interview with the Founder of Ekalavyas, Mr. Gopal, talking about the necessity of sports tech in India and its benefits.
Innosphere 23
I achieved 3rd place in the STEM category at Innosphere 23, Sigma Xi Conference. It was a great experience for me to talk about my project to a panel of AI and Computer Science experts. Have a look at my presentation!
Problem Statement, Motivation, and Aim
Being a national level basketball player, I've played with students from not just my school but across the country. Noticing many of them struggle with shot accuracy and following deep discussions with my basketball coach about the importance of shooting technique, I decided to build Sports SensAI. Sports SensAI is a sports tech app that uses the powers of AI to provide individualized feedback to players. It gives coaches access to improved feedback mechanisms and helps players visually analyze and improve their shooting skills, with the help of technology and individualised training.
Software and Hardware
Sports SensAI utilises computer-vision and pose-recognition to analyze various aspects of a player’s shooting motion. Anchored on a sturdy NVIDIA GeForce GTX 1050 Ti GPU, which easily supports the use of CUDA and cuDNN, this app offers statistical insights, visual deconstruction, and progress tracking via NumPy, SciPy, and Matplotlib. To account for unique athletic styles, I used OpenCV to split videos into frames and detect key data points on a player’s body. TensorFlow computed the points identified by OpenPose library, forming a skeletal framework for examination.
Trained with numerous basketball footage pulled from YouTube, it assessed shooting variables- release angle, knee angles, elbow position, and ball trajectory. Since then, I’ve collaborated with over a dozen sports academies and public schools in Delhi-NCR to employ my computationally-backed coaching method.
Sample Analysis and User Interface
Future Plans
Sports SensAI has not only aims to increase youth participation, but also lead to unprecedented success in regional to national tournaments. Sports in India faces the two-step hurdle of non-specialized training and insufficient resource allocation. Through Sports SensAI, I aim to apply unbiased predictive models to democratise personal sports training and analysis. I aspire to make India’s scoreboards dynamic and more inclusive.
In the future, I see Sports SensAI being used for many other sports too. Whether it's to analyze a cricketer's bowling action, a golfer's swing, or a footballer's body position while taking a penalty, there is immense potential that can be unlocked through Sports SensAI, and I intend on doing so!
Volunteering at Happy School
At The Happy School Gurugram, I have given basketball coaching to over 50 underprivileged students!
