Audio processing module

To use the Acoustic ID API for audio identification, the environment variable ACOUSTIC_ID_API_KEY needs to be set at .env. Register at the website to get a valid token.

Add to the .env file the following field:

ACOUSTIC_ID_API_KEY=*****

It's required to execute the following commands, for the installation of pydub (see NOTE 1) and essentia-tensorflow in order to run the audio processing service.

Download the following models files and place them in the suggested locations.

1. Download and place the following models under models/ directory:

• genre_rosamerica-vggish-audioset-1.pb

2. Download, rename the following model to audio_mdl.pth and place it under process_audio_thread/pretrained_models/ directory:

• audio_mdl.pth

Installation and launch

pip install -r requirements_process_audio.txt

python process_audio_thread/process_audio_thread.py

NOTES:

1- Install FFMPEG. pydub needs either sudo apt install ffmpeg (Linux) or brew install ffmpeg (Mac) in order to function correctly.