AI Voice Keyboard

A Windows application that converts voice to keyboard input using speech recognition technology.

Latest Update (v1.0.3)

• Fixed issues with recording state management
• Improved floating button synchronization
• Enhanced error handling and thread safety
• See full release notes

Features

• Two Recognition Modes:

  • Command Mode: Execute keyboard commands like "up", "down", "control c", etc.
  • String Mode: Type text as you speak naturally

• AI-Powered Recognition:

  • Uses Whisper.net for accurate speech recognition
  • Falls back to Windows Speech Recognition when needed

• Convenient Interface:

  • Floating overlay button for easy access
  • System tray integration
  • Minimalist design

Getting Started

Prerequisites

• Windows 10 or later
• .NET 8.0 Runtime
• Microphone for voice input

Installation

1. Download the latest release from the Releases page
2. Extract the ZIP file to a location of your choice
3. Run VoiceToKeyboard.exe

First Run

On first launch, AI Voice Keyboard will need to download the Whisper model. If it doesn't download automatically:

1. Download the model manually from: https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-medium.bin
2. Place the model file in the same directory as the application executable

Usage

Basic Controls

• Start/Stop Listening: Click the "Start Listening" button or use the floating button
• Switch Modes: Select Command or String mode using the radio buttons
• Floating Button: Enable/disable the floating overlay for easy access

Voice Commands

Command Mode

Say commands like:

• Navigation: "up", "down", "left", "right", "page up", "page down", "home", "end"
• Function keys: "f1", "f2", etc.
• Common shortcuts: "control c" (copy), "control v" (paste), "control z" (undo)
• Switch modes: "command option" or "string option"

String Mode

• Simply speak naturally and the system will type what you say
• Speech will automatically pause after 1.5 seconds of silence
• To stop recognition, click the recording button again

Troubleshooting

• No Speech Recognition: Ensure your microphone is properly connected and set as the default recording device
• Poor Recognition: Speak clearly and at a moderate pace
• Missing Whisper Model: Download the model file manually as described in the First Run section

License

This project is licensed under the GPL-3.0 License - see the LICENSE file for details.

Acknowledgments

• Whisper.net for AI speech recognition
• NAudio for audio processing
• Developed by ShioDev
• Powered by Beaver Foundation

Version

See VERSION.md for version history and release notes.