🤖 Human_Genx – Sketch-to-Face AI Generator

A cross-platform artistic AI application built using Flutter, Django, and a Pix2Pix deep learning model for transforming sketches into realistic human faces.

---

🧠 Overview

Human_Genx is a mobile application that leverages deep learning and modern app architecture to offer users a unique creative experience: draw a sketch, and receive a realistic human face.

This is achieved through the combination of:

• Flutter (Frontend UI)
• Firebase Authentication (Secure user login)
• Django Rest Framework (API & model serving)
• Pix2Pix model (Deep learning sketch-to-face mapping)

---

🔧 Tech Stack

Layer	Technology
Frontend	Flutter
Backend	Django REST Framework
Authentication	Firebase Auth
Deep Learning	TensorFlow + Keras (Pix2Pix model)
Model Deployment	Django API endpoint for image generation
Mobile Platforms	Android, iOS

---

📸 Project Preview

---

🎯 Key Features

• ✏️ Sketch-to-Real Human Face generation
• 🔐 Firebase Authentication
• 🌐 RESTful API communication (Flutter ↔ Django)
• 🤖 DL Model: Pix2Pix trained on facial data
• 🔁 Real-time image generation & response

---

🔄 Flow Summary

User Flow

[User draws sketch in app]
       ↓
[Sketch uploaded via Flutter]
       ↓
[Image sent to Django REST API]
       ↓
[Pix2Pix model generates realistic face]
       ↓
[Generated face sent back to user]

---

📐 System Architecture (Mermaid Diagram)

graph TD
    A[Flutter Frontend] --> B[Firebase Auth]
    A --> C[Sketch Upload]
    C --> D[Django API]
    D --> E[Pix2Pix DL Model]
    E --> F[Generated Image]
    F --> A

Loading

---

🧬 Sequence Diagram

sequenceDiagram
    participant U as User
    participant F as Flutter App
    participant B as Backend (Django)
    participant M as Pix2Pix Model

    U->>F: Logs in (Firebase)
    F->>B: Sends Sketch Image
    B->>M: Processes Sketch with Pix2Pix
    M-->>B: Returns Realistic Face
    B-->>F: Sends Image Response
    F-->>U: Displays Generated Image

Loading

---

🎭 Use Case Diagram

graph TD
    A[User] --> B[Login with Firebase]
    A --> C[Submit Sketch]
    A --> D[View Generated Face]
    B --> E[Flutter App]
    C --> E
    D --> E
    E --> F[Django Backend]
    F --> G[Pix2Pix Model]

Loading

---

🏁 Getting Started (For Developers)

📱 Flutter Setup

Make sure you have Flutter installed. Then:

git clone https://github.com/jayanth119/Human_Genx.git
cd Human_Genx/client
flutter pub get
flutter run

Update your Firebase configuration in:

• android/app/google-services.json
• ios/Runner/GoogleService-Info.plist

---

🧠 Backend Setup (Django + Pix2Pix)

cd Human_Genx/server
python -m venv venv
source venv/bin/activate  # or venv\Scripts\activate on Windows
pip install -r requirements.txt
python manage.py runserver

Make sure your model is trained and saved as pix2pix_model.h5 inside the model folder.

---

📂 Project Structure

Human_Genx/
├── client/                # Flutter app code
│   └── lib/
│       └── screens/
├── server/                # Django backend
│   ├── ml/                # Pix2Pix model code
│   ├── api/               # Django REST views & serializers
│   └── media/             # Uploaded & generated images
├── diagrams/              # All diagrams (UML, sequence, mermaid)
└── README.md              # Project documentation

---

🧪 Model Details – Pix2Pix

• Architecture: Conditional GAN
• Input: Binary sketch image (512x512)
• Output: RGB photo-realistic image
• Frameworks: TensorFlow 2.x, Keras
• Trained on: Custom dataset of faces + sketches

---

📈 Future Enhancements

• 📦 Deploy model to TensorFlow Serving or Hugging Face Inference API
• 🌍 Use WebSockets for real-time image feedback
• 🔄 Add Undo/Redo support for sketch input
• 📲 iOS and Android deployment on Play Store / App Store

---

✍️ Author

👤 Jayanth Chukka Full-stack Developer & AI Researcher 🔗 GitHub - jayanth119

---