audrey serves as a client to OpenAI Whisper, a general-purpose speech recognition model is trained on a large dataset of diverse audio. It can perform multilingual speech recognition, speech translation, and language identification.

You can install the development version of audrey from GitHub with:

# install.packages("devtools")
devtools::install_github("howardbaek/audrey")

Importantly, you need to download Whisper with the following command:

pip install -U openai-whisper

You also need the command-line tool ffmpeg on your system:

# on Ubuntu or Debian
sudo apt update && sudo apt install ffmpeg

# on Arch Linux
sudo pacman -S ffmpeg

# on MacOS using Homebrew (https://brew.sh/)
brew install ffmpeg

# on Windows using Chocolatey (https://chocolatey.org/)
choco install ffmpeg

# on Windows using Scoop (https://scoop.sh/)
scoop install ffmpeg

For more information, refer to the Whisper’s Setup documentation.

Convert the input audio file into written text using speech recognition, while preserving the language of the original audio:

library(audrey)

transcribe(system.file("extdata", "sample-english1.mp3",  package = "audrey"),
           output_dir = ".")

Provide an English translation of the input audio file that is originally spoken in a non-English language:

# Korean to English
translate(system.file("extdata", "sample-korean.wav",  package = "audrey"),
          audio_lang = "Korean", output_dir = ".")

• OpenAI Whisper model
• Source of audio files: https://commonvoice.mozilla.org/en/datasets