Android libraries
Improfit.ai
Built Android libraries providing real-time body-pose analysis and visual feedback for exercise and rehabilitation.
The project
I built an Android library for a gym equipment manufacturer to analyse body position and flexibility during exercises. It provided angles, repetition counts, instructions, constraints, and visual feedback. I also built reusable Android components for evaluating exercises defined in the company’s domain language.
Context
Improfit.ai is a company specialized in digitalizing movements of the skeleton for healthcare and fitness purposes. Improfit is helping to improve people's lives through training and rehabilitation using computer vision and artificial intelligence in real time.
My contribution
Integrated ML Kit and MediaPipe pose detection.
Integrated Intel RealSense to work with colour and depth frames.
Used OpenCV image preprocessing to improve pose detection in noisy environments.
Built Android libraries using the SDK and NDK, alongside sample apps.
At a glance
- Body movementPose & depth detection
- Android librariesExercise analysis
- Live feedbackAngles, repetitions & guidance
Toolkit
- Android