Dictate

A Linux-based voice dictation system using OpenAI's Whisper model for speech-to-text conversion and automatic text input.

Overview

This system provides real-time voice dictation capabilities by:

• Recording audio from your system's microphone
• Converting speech to text using Whisper
• Automatically typing the recognized text using ydotool

Prerequisites

• Linux system with systemd
• Python 3.x
• Root access for installation
• microphone
• ydotool installed

Installation

Either:

curl -sSL https://raw.githubusercontent.com/nilock/dictate/main/remote_install.sh | bash

Or:

1. Clone this repository:

git clone https://github.com/nilock/dictate
cd dictate

2. Run the installation script as root:

sudo ./installation.sh

This will:

• Create a Python virtual environment in /opt/dictation_venv
• Install required Python packages
• Set up the dictation service
• Start the service automatically

Usage

1. Start/stop dictation using the provided script:

dictation.sh

It is useful to set up a system hot-key to point to this script at its installation destination: /usr/local/bin/dictation.sh

2. When activated:

   • Speak
   • Run the command again to stop recording
   • The recognized text will be automatically typed at your cursor position

3. Check service status:

systemctl status dictation

4. View logs:

journalctl -u dictation -f

Components

• dictation_daemon.py: Background service handling audio recording and transcription
• dictation_client.py: Client interface for sending commands to the daemon
• dictation.sh: Convenient shell script wrapper
• installation.sh: System setup and service installation

Configuration

The system uses Whisper's "base" model by default. You can modify dictation_daemon.py to use different models:

Troubleshooting

1. Check service status:

systemctl status dictation

2. Review logs:

tail -f /tmp/dictation_daemon.log

3. Test audio recording:

python3 dictation_script_only.py

Debug Files

The system saves the last recording as /tmp/last_recording.wav for debugging purposes.

Development

A standalone testing script (dictation_script_only.py) is provided for development and testing purposes.

The following is useful to redeploy locally:

sudo bash ./installation.sh && sudo journalctl -u dictation -f

Dependencies

• openai-whisper
• sounddevice
• numpy
• torch
• scipy
• pynput

License

GPL-3.0

Contributing

Sure, go nuts.