Skip to content

Adalex3/PTAppHackUSF

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

111 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Physical Trainer Movement Helper App — Project Requirements

This project is a smart movement and rehabilitation assistant designed to guide users through physical exercises with live camera feedback. Users can either upload a treatment plan from their doctor or have GPT-4o generate a custom plan for them. The app leverages computer vision to identify form issues and provides detailed, real-time corrections through a sleek, easy-to-use interface.


Core Workflow

Step 1: Add Your Treatment Plan

  • Upload your prescribed exercise plan from a healthcare provider or
  • Use GPT-4o to generate a personalized plan based on your needs

Step 2: Perform the Exercises with Camera Feedback

  • The app uses Mediapipe and OpenCV for pose estimation
  • You’ll be guided through hardcoded exercises that we’ve trained the app to understand
  • Real-time feedback will tell you what’s wrong and how to fix it, with actionable and specific suggestions

Step 3: Get a Visual Summary

  • After you’re done, you’ll see a clean, visual breakdown of your performance
  • Charts, highlights, and annotations will show what you did well and what needs improvement

Optional Features

Mistake Highlight Reel

  • The app can clip short video segments where your form broke down
  • Playback lets you review mistakes with visual overlays and suggestions

Doctor Review Portal

  • Option to send your feedback and performance to your doctor
  • Uses Firebase to store and sync data
  • Doctors can log in through a separate interface to check in on your progress

Phase 2: Raspberry Pi Integration (Optional)

In the future, we’ll allow you to run the pose estimation on a Raspberry Pi with a camera module. It’ll send pose data over WiFi to your computer, offloading the heavy lifting and letting the app run more smoothly on lightweight systems.


Tech Stack

  • Frontend: React, Tailwind CSS, Recharts or Chart.js
  • Pose Estimation: Mediapipe, OpenCV (Python)
  • Backend: Firebase (for auth and storage), Flask or FastAPI (for GPT integration)
  • Video Handling: FFmpeg or MediaRecorder API
  • Optional Hardware: Raspberry Pi 4, Camera Module v2

Milestones

  1. Treatment Plan UI + GPT generation
  2. Pose Detection + Feedback for Core Movements
  3. Real-Time Guidance with Specific Corrections
  4. Visual Summary Dashboard
  5. Mistake Replay System (optional)
  6. Doctor Sharing Portal (optional)
  7. Raspberry Pi Streaming Support (optional)

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •