An implementation of music separation model by Luo et.al.

0. Prepare .wav files to separate.

1. Install library pip install git+https://github.com/leichtrhino/ChimeraNet

2. Download pretrained model.

3. Download sample script.

4. Run script

python chimeranet-separate.py -i ${input_dir}/*.wav \
    -m model.hdf5 \
    --replace-top-directory ${output_dir}

Output in nutshell

• the format of filename is ${input_file}_{embd,mask}_ch[12].wav.
• embd and mask indicates that it was inferred from deep clustering and mask respectively.
• ch1 and ch2 are voice and music channel respectively.

See Example section on ChimeraNet documentation.

• keras
• one of keras' backends (i.e. TensorFlow, CNTK, Theano)
• sklearn
• librosa
• soundfile

1. Run pip install git+https://github.com/leichtrhino/ChimeraNet or any python package installer. (Currently, ChimeraNet is not in PyPI.)
2. Install keras' backend if the environment does not have any. Install tensorflow if unsure.

• Luo et.al.
• Hershey et.al.
• Original ChimeraNet site