✍️ Multilingual Handwriting AI

A beautiful, full-screen web application that converts handwriting into digital text in real-time. Built with Material 3 Expressive Design, it supports 20+ languages including Hindi, English, Greek, Chinese, and more.

🚀 Live Demo

Click here to use the App

✨ Features

• Real-time Recognition: Instantly converts strokes to text using Google's Input Tools API.
• Zero Latency: Immediate processing upon lifting the mouse/finger (no input lag).
• Material 3 UI: Modern, "Expressive" design using the Violet/Purple tonal palette.
• Smart Layout:
  • Fixed Sidebar: The interface never jumps or shrinks while you write.
  • 3-Card Suggestions: Displays the Top 3 results in a static grid (Left, Center-Highlighted, Right).
• Cross-Platform:
  • Desktop: Full-screen whiteboard experience.
  • Mobile: Responsive vertical layout for writing on the go.
• Searchable Language List: Easily find your language with a built-in search bar.
• Utility Tools: One-click "Copy to Clipboard" and "Clear Text" functions.

🌍 Supported Languages

Supports over 20+ languages, including but not limited to:

• Hindi (हिन्दी)
• English
• Greek (Ελληνικά)
• Chinese (中文)
• Japanese (日本語)
• Russian (Русский)
• Arabic (العربية)
• Spanish, French, German, Italian, Portuguese
• Bengali, Marathi, Tamil, Telugu, Kannada, Malayalam, Punjabi, Urdu, Nepali
• and many more.

🛠️ How to Run Locally

This project is built as a Single Page Application (SPA) with zero dependencies. You do not need Node.js, React, or Python to run it.

1. Clone the repository:

   git clone github.com:Kuldeep-Dilliwar/Multilingual-Handwriting-AI.git

2. Navigate to the folder.
3. Double-click index.html to open it in your browser.

Note: An internet connection is required for the handwriting recognition to work (it connects to Google's API).

🧩 How it Works

1. Canvas API: Captures mouse and touch coordinates (X, Y, Time) as "Strokes".
2. Fetch API: Sends these stroke arrays to the Google Input Tools endpoint (inputtools.google.com).
3. JSON Processing: Parses the response and extracts the top 3 predicted words.
4. DOM Manipulation: Updates the suggestion cards and text area dynamically.

⚠️ API Disclaimer

This project utilizes the undocumented public API endpoints used by Google Input Tools.

• No API Key required.
• Free for personal and educational use.
• Data is processed by Google's servers. Do not use for highly sensitive/confidential data.

📸 Screenshots

📄 License

This project is open source and available under the MIT License.

---

Made with ❤️ by [Kuldeep Dilliwars]