aiTools — small audio / assistant utilities

Brief descriptions of the subprojects in this workspace:

• assistent: A small assistant application that provides keyboard-driven automation and voice-related helper functions. See assistent.
• chatterbox: Tools for generating long German voice output; includes scripts to synthesize and play speech. See chatterbox.
• stt: Speech To Text / Push To Talk utilities including a live hotkey-driven recorder that transcribes audio using Whisper. It inserts the transcription in the current active window. See stt.

Quick start

Each subproject has its own requirements and usage notes. Enter the subproject directory and follow the README there, for example:

cd stt
source .venv/bin/activate
pip install -r requirements.txt
python3 src/speechToText.py

Notes

• On Ubuntu you may need system packages for audio and media handling, for example: libportaudio2, portaudio19-dev, libasound2-dev, and ffmpeg.
• GPU acceleration for Whisper/torch requires a CUDA-compatible PyTorch build; see the project READMEs for details.

License

See the LICENSE files in each subproject.