script for extracting multiple audio representations.
Setup environment with conda:
conda env create -f extract_env.yml
conda activate extract
Run the script on all audio files in a directory
python extract_features.py /path/to/input/audio/directory /path/to/feature/output/directory [--recompute]
Use the optional flag --recompute
to force feature computation on all files in the audio directory.
Otherwise, skip existing files in the feature output directory.