def music_mix_and_sources(musdb_tracks):
dataset = datasets.MUSDB18(folder=musdb_tracks.root,
download=False,
transform=transforms.SumSources(
[['drums', 'bass', 'other']]))
item = dataset[0]
return item['mix'], item['sources']
def test_transform_sum_sources(musdb_tracks):
track = musdb_tracks[10]
mix, sources = nussl.utils.musdb_track_to_audio_signals(track)
data = {'mix': mix, 'sources': sources}
groups = itertools.combinations(data['sources'].keys(), 3)
tfm = None
for group in groups:
_data = copy.deepcopy(data)
tfm = transforms.SumSources([group])
_data = tfm(_data)
for g in group:
assert g not in _data['sources']
assert '+'.join(group) in _data['sources']
summed_sources = sum([sources[k] for k in group])
assert np.allclose(_data['sources']['+'.join(group)].audio_data,
summed_sources.audio_data)
pytest.raises(TransformException, tfm, {'no_key'})
pytest.raises(TransformException, transforms.SumSources, 'test')
pytest.raises(TransformException, transforms.SumSources,
[['vocals', 'test'], ['test2', 'test3']], ['mygroup'])
def transform(stft_params: nussl.STFTParams,
sample_rate: int,
target_instrument,
only_audio_signal: bool,
mask_type: str = 'msa',
audio_only: bool = False):
"""
Builds transforms that get applied to
training and validation datasets.
Parameters
----------
stft_params : nussl.STFTParams
Parameters of STFT (see: signal).
sample_rate : int
Sample rate of audio signal (see: signal).
target_instrument : str
Which instrument to learn to separate out of
a mixture.
only_audio_signal : bool
Whether to return only the audio signals, no
tensors (useful for eval).
mask_type : str, optional
What type of masking to use. Either phase
sensitive spectrum approx. (psa) or
magnitude spectrum approx (msa), by default
'msa'.
audio_only : bool, optional
Whether or not to only apply GetAudio in
transform (don't compute STFTs).
"""
tfm = []
other_labels = [k for k in LABELS if k != target_instrument]
tfm.append(nussl_tfm.SumSources([other_labels]))
new_labels = [target_instrument] + tfm[-1].group_names
new_labels = sorted(new_labels)
if not only_audio_signal:
if not audio_only:
if mask_type == 'psa':
tfm.append(nussl_tfm.PhaseSensitiveSpectrumApproximation())
elif mask_type == 'msa':
tfm.append(nussl_tfm.MagnitudeSpectrumApproximation())
tfm.append(nussl_tfm.MagnitudeWeights())
tfm.append(nussl_tfm.GetAudio())
target_index = new_labels.index(target_instrument)
tfm.append(nussl_tfm.IndexSources('source_magnitudes', target_index))
tfm.append(nussl_tfm.ToSeparationModel())
return nussl_tfm.Compose(tfm), new_labels
def test_transform_compose(musdb_tracks):
track = musdb_tracks[10]
mix, sources = nussl.utils.musdb_track_to_audio_signals(track)
data = {
'mix': mix,
'sources': sources,
'metadata': {
'labels': ['bass', 'drums', 'other', 'vocals']
}
}
class _BadTransform(object):
def __call__(self, data):
return 'not a dictionary'
com = transforms.Compose([_BadTransform()])
pytest.raises(TransformException, com, data)
msa = transforms.MagnitudeSpectrumApproximation()
tfm = transforms.SumSources(
[['other', 'drums', 'bass']],
group_names=['accompaniment']
)
assert isinstance(str(tfm), str)
com = transforms.Compose([tfm, msa])
assert msa.__class__.__name__ in str(com)
assert tfm.__class__.__name__ in str(com)
data = com(data)
assert np.allclose(data['mix_magnitude'], np.abs(mix.stft()))
assert data['metadata']['labels'] == [
'bass', 'drums', 'other', 'vocals', 'accompaniment']
mask_data = (
data['source_magnitudes'] /
np.maximum(
data['mix_magnitude'][..., None],
data['source_magnitudes'])
)
msa_scores = separate_and_evaluate(mix, data['sources'], mask_data)
shape = mix.stft_data.shape + (len(sources),)
mask_data = np.ones(shape)
mix_scores = separate_and_evaluate(mix, data['sources'], mask_data)
for key in msa_scores:
if key in ['SI-SDR', 'SI-SIR', 'SI-SAR']:
diff = np.array(msa_scores[key]) - np.array(mix_scores[key])
assert diff.mean() > 10
def test_dataset_hook_scaper_folder(scaper_folder):
dataset = nussl.datasets.Scaper(scaper_folder)
data = dataset[0]
_sources = [data['sources'][k] for k in data['sources']]
assert np.allclose(sum(_sources).audio_data, data['mix'].audio_data)
for k in data['sources']:
assert k.split('::')[0] in data['metadata']['labels']
# make sure SumSources transform works
tfm = transforms.SumSources(
[['050', '051']],
group_names=['both'],
)
data = tfm(data)
for k in data['sources']:
assert k.split('::')[0] in data['metadata']['labels']
_sources = [data['sources'][k] for k in data['sources']]
assert np.allclose(sum(_sources).audio_data, data['mix'].audio_data)