Ensemble Clustering

Seetharaman, Prem. Bootstrapping the Learning Process for Computer Audition. Diss. Northwestern University, 2019.

@phdthesis{seetharaman2019bootstrapping,
  title={Bootstrapping the Learning Process for Computer Audition},
  author={Seetharaman, Prem},
  year={2019},
  school={Northwestern University}
}

import nussl
import matplotlib.pyplot as plt
import time
import warnings
import numpy as np

warnings.filterwarnings("ignore")
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=-40, alpha_amount=2.0)
    plt.subplot(212)
    nussl.utils.visualize_sources_as_waveform(
        sources, show_legend=False)
    plt.show()
    nussl.play_utils.multitrack(sources)

audio_path = nussl.efz_utils.download_audio_file(
    'schoolboy_fascination_excerpt.wav')
audio_signal = nussl.AudioSignal(audio_path)

separators = [
    nussl.separation.primitive.FT2D(audio_signal),
    nussl.separation.primitive.HPSS(audio_signal),
    nussl.separation.primitive.Melodia(audio_signal),
]

weights = [2, 1, 2]
returns = [[1], [0], [1]]

fixed_centers = np.array([
    [0 for i in range(sum(weights))],
    [1 for i in range(sum(weights))],
])

ensemble = nussl.separation.composite.EnsembleClustering(
    audio_signal, 2, separators=separators, init=fixed_centers,
    fit_clusterer=False, weights=weights, returns=returns)
ensemble.clusterer.cluster_centers_ = fixed_centers
estimates = ensemble()

estimates = {
    f'Cluster {i}': e for i, e in enumerate(estimates)
}

visualize_and_embed(estimates)

Matching file found at /home/pseetharaman/.nussl/audio/schoolboy_fascination_excerpt.wav, skipping download.

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

Time taken: 19.5243 seconds