Whisper GUI

A lightweight alternative to Windows H (Windows Speech Recognition) that provides more flexibility and language support using OpenAI's Whisper model.

Overview

This tool is designed to replace the built-in Windows H speech recognition feature, offering enhanced capabilities for multilingual speech-to-text transcription. It's particularly useful for users who need to transcribe speech in multiple languages (English, Spanish, etc.) with high accuracy.

Features

• 🎤 Simple one-click recording interface
• 🌍 Multi-language support (English, Spanish, and more)
• 📝 Real-time transcription
• 🎯 High accuracy using OpenAI's Whisper model
• 🖥️ Clean, simple interface

Demo

Here's a screenshot of the application interface:

Installation

Option 1: Download Pre-built Executable

1. Go to the Releases page
2. Download the latest version of whisper-gui.exe
3. Run the executable

Option 2: Build from Source

1. Clone the repository:

   git clone https://github.com/elpargo/whisper-windows-gui/releases
   cd whisper-gui

2. Create and activate a virtual environment:

   python -m venv venv
   .\venv\Scripts\activate

3. Install dependencies:

   pip install -r requirements.txt

4. Build the executable:

   .\build.ps1

Usage

1. Run the application
2. Click the microphone button (or press Space/Enter) to start recording
3. Speak in your desired language
4. Click the button again (or press Space/Enter) to stop recording
5. The transcription will appear in the text area
6. Click "Save" to save the transcription to a file

Note: You can use either the Space or Enter key interchangeably to start/stop recording.

Why Whisper GUI?

While Windows H provides basic speech recognition, it has limitations:

• Limited language support
• Requires internet connection
• Less accurate for non-English languages
• No easy way to save transcriptions

Whisper GUI addresses these issues by:

• Supporting multiple languages
• Working offline
• Providing higher accuracy
• Offering easy transcription saving

Technical Details

• Built with Python and PyQt6
• Uses OpenAI's Whisper model for transcription
• Compiled with PyInstaller for easy distribution

License

This project is released under the MIT License. See the LICENSE file for details.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Acknowledgments

• OpenAI Whisper for the speech recognition model
• PyQt6 for the GUI framework

TODOs

• 💾 Save transcriptions to text files
• Invoke on global OS keybinding (ie: replace windows + H entirely)
• output to text input field directly (Also a windows + H feature)