HPSS

Fitzgerald, Derry. “Harmonic/percussive separation using median filtering.” 13th International Conference on Digital Audio Effects (DAFX10), Graz, Austria, 2010.

Driedger, Müller, Disch. “Extending harmonic-percussive separation of audio.” 15th International Society for Music Information Retrieval Conference (ISMIR 2014) Taipei, Taiwan, 2014.

@inproceedings{fitzgerald2010harmonic,
  title={Harmonic/percussive separation using median filtering},
  author={Fitzgerald, Derry},
  booktitle={Proc. of DAFX},
  volume={10},
  number={4},
  year={2010}
}

@inproceedings{driedger2014extending,
  title={Extending Harmonic-Percussive Separation of Audio Signals.},
  author={Driedger, Jonathan and M{\"u}ller, Meinard and Disch, Sascha},
  booktitle={ISMIR},
  pages={611--616},
  year={2014}
}

import nussl
import matplotlib.pyplot as plt
import time

start_time = time.time()

def visualize_and_embed(sources):
    plt.figure(figsize=(10, 6))
    plt.subplot(211)
    nussl.utils.visualize_sources_as_masks(sources,
        y_axis='mel', db_cutoff=-60, alpha_amount=2.0)
    plt.subplot(212)
    nussl.utils.visualize_sources_as_waveform(
        sources, show_legend=False)
    plt.show()
    nussl.play_utils.multitrack(sources)

tfm = nussl.datasets.transforms.SumSources([
    ['vocals', 'other', 'bass']
])
musdb = nussl.datasets.MUSDB18(
    download=True, transform=tfm)
i = 40

item = musdb[i]
mix = item['mix']
sources = item['sources']

visualize_and_embed(sources)

separator = nussl.separation.primitive.HPSS(
    mix, mask_type='binary')
estimates = separator()
estimates = {
    'Harmonic': estimates[0],
    'Percussive': estimates[1]
}

visualize_and_embed(estimates)

end_time = time.time()
time_taken = end_time - start_time
print(f'Time taken: {time_taken:.4f} seconds')

Time taken: 20.9481 seconds