HandwritingGen

HandwritingGen is an application for synthesizing handwriting from a user's training input data. The app lets you train a model on user-provided handwriting samples and then generate new handwriting images or stroke sequences in the learned style. The project separates concerns into a Python backend (model training / inference) and a JavaScript frontend (UI for uploading samples, starting training, previewing synthesized handwriting).

---

Key Features

• Train a handwriting synthesis model on user-provided samples
• Serve model inference via an HTTP API
• Web UI for uploading training data, monitoring training, and generating samples

---

Target Users

• Programmers and Data Science researchers and students who want a quick pipeline for handwriting synthesis experiments
• Developers building handwriting-style features (font generation, stylized text)
• Hobbyists interested in generative handwriting models

---

Tech Stack

• Backend: Python (FastAPI recommended) for model training, inference, and REST API with Uvicorn
• Model code: Numpy to store and cross-compute across data
• Frontend: React (create-react-app / Vite) with Babel
• Development tooling: npm / npx for frontend utilities, Python venv for backend isolation

---

Prerequisites

• Python 3 (3.13 recommended)
• pip
• Node.js 16+ and npm (npx is included with npm)
• A local IP address for cross-device testing (if you want to access the backend from another machine / phone)
• (Optional) Expo Go mobile app (for running app on phone or device other than web)

---

Environment Variables

You must configure an environment variable for the frontend to connect to the proper IP address. Create the file .env in the frontend/ directory, and save the IP address on which you will run the server. This could be localhost or your local IP in the form of EXPO_PUBLIC_API_URL=http://<IPaddress>:<port>.

Example frontend .env with default, local frontend connection configuration:

EXPO_PUBLIC_API_URL=http://localhost:5000

Important: Replace localhost with the IP address of the machine running your backend if you want to access it from other devices on the same LAN. Passing 0.0.0.0 to the backend host exposes it to the network (uvicorn / FastAPI CLI uses --host 0.0.0.0), but API clients need the concrete IP (not 0.0.0.0) to call from outside. To get yours, enter the ipconfig command into the terminal and look for IPv4 Address.

---

Setup (step-by-step)

Below are typical steps to prepare and run both backend and frontend locally.

1) Clone the repo

git clone https://github.com/ajfrist/HandwritingGen.git
cd HandwritingGen

2) Backend (Python) setup

1. Create and activate a virtual environment

• Linux / macOS:

  python3 -m venv venv
  source venv/bin/activate

• Windows (PowerShell):

  python -m venv venv
  .\venv\Scripts\Activate.ps1

2. Install Python packages

# from repository root, with venv active
python -m pip install --upgrade pip
pip install -r python/requirements.txt

3. Run the backend (FastAPI / Uvicorn example)

# from repository root
cd python
# ensure venv is active
python -m uvicorn api_server:app --port 5000 --reload

To expose backend to devices on the LAN network, insead run:

python -m uvicorn api_server:app --host 0.0.0.0 --port 5000 --reload

3) Frontend (Node) setup

1. Install dependencies

# from repository root
cd frontend
npm install

2. Configure frontend .env (see above Environment Variables):

3. Run the frontend (development)

npx expo start

This will display directions in the terminal for running the application on the web or in emulator. To run on a physical mobile device, scan the QR code with the Expo Go app.

4) Mobile Application Connection

1. Download Expo Go

• Install Expo Go from your mobile app store.

2. Scan QR Code

• Open the Expo Go app and scan the QR code generated in the terminal.
• This will build the application for your device and allow you to use it on external hardware.