🍜 Vietnamese Food Detection

AI-powered mobile app for detecting Vietnamese dishes with nutritional information.

📊 Project Architecture

graph TB
    subgraph "Mobile App (React Native)"
        A[📱 iOS/Android App] --> B[Camera/Gallery]
        B --> C[Image Capture]
        C --> D[API Client]
    end
    
    subgraph "Backend (FastAPI)"
        D --> E[🚀 Railway Server]
        E --> F[EfficientNet B4]
        F --> G[Classification]
        G --> H[Nutrition Lookup]
    end
    
    subgraph "ML Training"
        I[📊 Dataset] --> J[Training Pipeline]
        J --> K[YOLOv10 + EfficientNet]
        K --> L[Trained Models]
        L --> F
    end
    
    H --> M[📱 Results Display]
    M --> N[Share Results]
    
    style A fill:#4CAF50
    style E fill:#FF6B6B
    style K fill:#4ECDC4

Loading

🌿 Branch Structure

vn-food-detection/
│
├── 📱 mobile-app          # React Native + Backend
│   ├── VNFoodDetection/   # Mobile app
│   └── server/            # FastAPI + Model
│
└── 🧠 training            # ML Training Pipeline
    ├── src/               # Training scripts
    ├── data_master/       # Dataset (30 classes)
    └── models/            # Trained models

🚀 Quick Start

Mobile App Development

git clone -b mobile-app https://github.com/silent9669/vn-food-detection.git
cd vn-food-detection/mobile-app

ML Training

git clone -b training https://github.com/silent9669/vn-food-detection.git
cd vn-food-detection
./menu.sh

📱 Mobile App Features

flowchart LR
    A[🏠 Home] --> B{Choose Input}
    B -->|Camera| C[📸 Capture]
    B -->|Gallery| D[🖼️ Select]
    C --> E[🔄 Processing]
    D --> E
    E --> F[✅ Results]
    F --> G[📊 Nutrition Info]
    F --> H[📤 Share]
    
    style A fill:#E3F2FD
    style F fill:#C8E6C9
    style G fill:#FFF9C4

Loading

Tech Stack:

• React Native + TypeScript
• Platform-specific UI (iOS/Android)
• 69+ tests (92% pass rate)
• Railway backend deployment

🧠 ML Pipeline

flowchart TD
    A[Raw Images] --> B[Data Augmentation]
    B --> C{Training}
    C -->|Classification| D[EfficientNet B4]
    C -->|Detection| E[YOLOv10]
    D --> F[97%+ Accuracy]
    E --> F
    F --> G[Production Model]
    G --> H[Mobile Backend]
    
    style A fill:#FFE0B2
    style F fill:#C8E6C9
    style H fill:#B2DFDB

Loading

Capabilities:

• 30 Vietnamese food classes
• 17,581 training images
• Hybrid detection system
• Nutrition calculation

📊 System Flow

sequenceDiagram
    participant U as 👤 User
    participant M as 📱 Mobile App
    participant B as 🚀 Backend
    participant AI as 🤖 AI Model
    
    U->>M: Take/Select Photo
    M->>B: Upload Image
    B->>AI: Process Image
    AI->>AI: EfficientNet Classification
    AI->>B: Detection Results
    B->>B: Calculate Nutrition
    B->>M: Return Results
    M->>U: Display Results + Nutrition
    U->>M: Share Results

Loading

🎯 Supported Dishes (30 Classes)

Category	Examples
Noodles	Phở, Bún bò Huế, Bún chả
Rice	Cơm tấm, Cơm gà
Bread	Bánh mì
Cakes	Bánh xèo, Bánh cuốn, Bánh bèo
Rolls	Gỏi cuốn, Nem chua
Others	20+ more dishes

🔗 Links

• Mobile App Branch: mobile-app
• Training Branch: training
• Documentation: See branch-specific docs

📞 Contact

Email: phuc.dangcs2007@hcmut.edu.vn

---

🍜 Detect • Analyze • Track
Vietnamese Food Detection with AI