Inspiration

Talent is everywhere, but opportunities are not. Many students, especially in rural areas, lack access to professional coaches, sports facilities, and proper talent assessment. We wanted to make sports talent discovery accessible through a smartphone.

What it does

Skill2Sport AI analyzes videos of users performing 5 simple sports tests such as jumping, speed, and agility. AI evaluates each test, generates scores, calculates sports-wise readiness percentages, and provides personalized sport recommendations.

How we built it

We use React Native/Flutter for the app, MediaPipe and OpenCV for pose and video analysis, PyTorch/TensorFlow for AI/ML, and FastAPI/Node.js with PostgreSQL for the backend and data management.

Challenges we ran into

Handling different camera angles, lighting, video quality, and recording environments was challenging. Converting body movements into accurate and meaningful scores was another major challenge.

Accomplishments that we're proud of

We created an end-to-end concept that transforms simple videos into performance scores, sports readiness percentages, and personalized recommendations.

What we learned

We learned how computer vision, AI, video processing, and backend technologies can work together to solve real-world problems. We also learned the importance of data quality and real-world testing.

What's next for Skill2Sport AI

We plan to add more sports tests, improve AI accuracy, enable real-time mobile analysis, and build dashboards for coaches, schools, and scouts, making talent discovery more accessible nationwide.

Built With

Share this project:

Updates

Submission history