Task
Teach a basketball hoop to see the player: record shots and hits so that solo training gives the same feedback as a session with a coach.
Recognition had to work on the device itself — fast and stable in any room and under any lighting.
US startup
AI basketball hoop
The client is a US startup.
Computer visionEdge inferenceEmbedded systemsSports analytics
Computer vision for a smart basketball hoop with a screen : shot recognition and player statistics.
Client
US startup
Industry
Sport
Areas
Computer vision, embedded systems
ESTESIS role
Technology partner
From task to result
Client
A US startup producing a smart basketball hoop with a built-in screen and camera. We do not disclose the client's name.
What we did and how
We developed computer vision modules that run on the device and feed data into the game and training modes.
- 01Ball detection and recording of hits and misses
- 02Detecting the player's position and shot point
- 03Training statistics for the player
- 04Optimizing models for the device's computing resources
Result
The hoop itself recognizes shots and keeps player statistics, and training becomes an interactive format with clear feedback.
Stack and competencies
More stories
Your task —next story
Tell us about the project — within 5 working days we will send an architecture and an estimate.
Accredited IT company · Russian Ministry of Digital Development registry No. 20142