🚀 FastAPI + Rust High-Performance Edge Computing

EdgeFit-AR 🥽 Intelligent Glasses Sports Assistance System, achieving millisecond-level feedback for sports injury prevention on smart glasses.

✨ Project Features

• ⚡ 5ms Level Real-Time Response - High-performance pose analysis algorithm implemented in Rust
• 🧠 AI Intelligent Error Correction - PyTorch LSTM model for real-time detection of motion errors
• 👓 AR Visualization - Support for Web/Unity dual interfaces
• 🔋 Edge Computing - Local inference, protecting user privacy
• 📊 Data Source - Supports UCI HAR fitness data (expandable in the future)

🚀 Quick Start

Environment Requirements

• Python 3.11
• Rust 1.68+
• 8GB+ RAM

📁 Project Structure

EdgeFit-AR/
├── 📊 data_engine/                # Data Processing Core
│   ├── datasets/                   # Dataset Storage
│   │   ├── raw/                    # Raw Data (UCI HAR)
│   │   └── processed/              # Processed Data
│   ├── preprocessing.py            # Data Preprocessing
│   └── setup_datasets.py           # Dataset Configuration
│
├── ⚡ edge_gateway/               # Edge Computing Gateway
│   ├── api/                        # FastAPI
│   │   ├── main.py                 # Service Entry
│   │   ├── model_manager.py        # AI Model Management
│   │   ├── connection_manager.py   # Conncetion
│   │   └── data_adapter.py         # Data Format Conversion
│   └── rust_engine/                # Rust High-Performance Engine 
│       ├── Cargo.toml              # Rust project configuration
│       ├── pyproject.toml          # Python build configuration
│       ├── build.py                # Maturin build
│       └── src/
│            ├── lib.rs             # Rust core implementation            
│            ├── features.rs        # Feature management code
│            ├── health.py          # Health check functionality
│            ├── pose.py            # Pose detection logic
│            └── inference.py       # Inference processing code     
│
├── 🥽 ar_interface/               # AR User Interface
│   └── web_simulator.py            # Web AR Simulator
│
├── 🤖 training/                   # AI Model Training
│   └── train_model.py              # Training Main Script

├── ⚙️ config/                     # Configuration Files (Auto-Generated)
└── 📋 requirements.txt            # Python Dependencies

🎯 Core Functions

1. Real-Time Pose Analysis

• LSTM neural network-based action recognition
• Supports squats, lunges, and other exercise types
• Millisecond-level error detection and feedback

2. Multi-Modal Data Fusion

• 6-axis IMU sensor data (accelerometer + gyroscope)
• Real-time data preprocessing and feature extraction
• Adaptive noise filtering

3. AR Visualization Interface

• Web version: Ready to use, no additional software needed
• Unity version: Full 3D AR experience (planned)
• Real-time pose correction guidance

🛠️ Development Roadmap

• v0.1 - Basic data processing and AI training
• v0.2 - Web AR interface and real-time inference
• v0.3 - Rust performance optimization engine
• v0.4 - Unity 3D AR interface
• v0.5 - Mobile deployment support
• v0.6 - Cloud synchronization and multi-user support

🤝 Contribution Guidelines

Welcome to contribute code! Please follow these steps:

1. Fork this project
2. Create a feature branch (git checkout -b feature/AmazingFeature)
3. Commit changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Create a Pull Request

📄 Dataset Support

UCI HAR Dataset - Placed in data_engine/datasets/raw/uci_har/

📞 Contact Information

• Author: Rafa-Gu98
• Email: rafagr98.dev@gmail.com
• Project Homepage: https://github.com/Rafa-Gu98/edgeFit-ar

📜 Open Source License

This project uses the MIT open source license.

🙏 Acknowledgments

• UCI Machine Learning Repository - HAR Dataset
• PyTorch - ML
• FastAPI - Web
• Rust - Core+

---

⭐ If this project helps you, please give it a Star!