def test_apply_to_all_cuda(self):
samples = np.array(
[
[[0.75, 0.5, -0.25, -0.125, 0.0]],
[[0.9, 0.5, -0.25, -0.125, 0.0]],
[[0.9, 0.5, -0.25, -1.12, 0.0]],
],
dtype=np.float32,
)
sample_rate = 16000
augment = PeakNormalization(p=1.0, output_type="dict")
processed_samples = (
augment(samples=torch.from_numpy(samples).cuda(), sample_rate=sample_rate)
.samples.cpu()
.numpy()
)
assert_almost_equal(
processed_samples,
np.array(
[
[[0.75 / 0.75, 0.5 / 0.75, -0.25 / 0.75, -0.125 / 0.75, 0.0 / 0.75]],
[[0.9 / 0.9, 0.5 / 0.9, -0.25 / 0.9, -0.125 / 0.9, 0.0 / 0.9]],
[[0.9 / 1.12, 0.5 / 1.12, -0.25 / 1.12, -1.12 / 1.12, 0.0 / 1.12]],
],
dtype=np.float32,
),
)
self.assertEqual(processed_samples.dtype, np.float32)
def test_apply_to_only_too_loud_sounds(self):
samples = np.array(
[
[[0.75, 0.5, -0.25, -0.125, 0.0]],
[[0.9, 0.5, -0.25, -0.125, 0.0]],
[[0.9, 0.5, -0.25, -1.12, 0.0]],
],
dtype=np.float32,
)
sample_rate = 16000
augment = PeakNormalization(
apply_to="only_too_loud_sounds", p=1.0, output_type="dict"
)
processed_samples = augment(
samples=torch.from_numpy(samples), sample_rate=sample_rate
).samples.numpy()
assert_almost_equal(
processed_samples,
np.array(
[
[[0.75, 0.5, -0.25, -0.125, 0.0]],
[[0.9, 0.5, -0.25, -0.125, 0.0]],
[[0.9 / 1.12, 0.5 / 1.12, -0.25 / 1.12, -1.12 / 1.12, 0.0 / 1.12]],
],
dtype=np.float32,
),
)
self.assertEqual(processed_samples.dtype, np.float32)
def test_never_apply(self):
samples = np.array(
[
[[0.75, 0.5, -0.25, -0.125, 0.0]],
[[0.9, 0.5, -0.25, -0.125, 0.0]],
[[0.9, 0.5, -0.25, -1.12, 0.0]],
],
dtype=np.float32,
)
sample_rate = 16000
augment = PeakNormalization(p=0.0, output_type="dict")
processed_samples = augment(
samples=torch.from_numpy(samples), sample_rate=sample_rate
).samples.numpy()
assert_equal(
processed_samples,
np.array(
[
[[0.75, 0.5, -0.25, -0.125, 0.0]],
[[0.9, 0.5, -0.25, -0.125, 0.0]],
[[0.9, 0.5, -0.25, -1.12, 0.0]],
],
dtype=np.float32,
),
)
self.assertEqual(processed_samples.dtype, np.float32)
def test_apply_to_all_2_dim(self):
samples = np.array(
[
[0.75, 0.5, -0.25, -0.125, 0.0],
[0.9, 0.5, -0.25, -0.125, 0.0],
[0.9, 0.5, -0.25, -1.12, 0.0],
],
dtype=np.float32,
)
sample_rate = 16000
augment = PeakNormalization(p=1.0)
processed_samples = augment(
samples=torch.from_numpy(samples), sample_rate=sample_rate
).numpy()
assert_almost_equal(
processed_samples,
np.array(
[
[0.75 / 0.75, 0.5 / 0.75, -0.25 / 0.75, -0.125 / 0.75, 0.0 / 0.75],
[0.9 / 0.9, 0.5 / 0.9, -0.25 / 0.9, -0.125 / 0.9, 0.0 / 0.9],
[0.9 / 1.12, 0.5 / 1.12, -0.25 / 1.12, -1.12 / 1.12, 0.0 / 1.12],
],
dtype=np.float32,
),
)
self.assertEqual(processed_samples.dtype, np.float32)
def test_digital_silence_in_batch(self):
"""Check that there is no division by zero in case of digital silence (all zeros)."""
samples = np.array(
[[[0.75, 0.5, -0.25, -0.125, 0.0]], [[0.0, 0.0, 0.0, 0.0, 0.0]]],
dtype=np.float32,
)
sample_rate = 16000
augment = PeakNormalization(p=1.0)
processed_samples = augment(samples=torch.from_numpy(samples),
sample_rate=sample_rate).numpy()
assert_almost_equal(
processed_samples,
np.array(
[
[[
0.75 / 0.75, 0.5 / 0.75, -0.25 / 0.75, -0.125 / 0.75,
0.0 / 0.75
]],
[[0.0, 0.0, 0.0, 0.0, 0.0]],
],
dtype=np.float32,
),
)
self.assertEqual(processed_samples.dtype, np.float32)
def test_stereo_sound(self):
samples = np.array(
[
[[0.75, 0.5, -0.25, -0.125, 0.0],
[0.65, 0.5, -0.25, -0.125, 0.0]],
[[0.3, 0.5, -0.25, -0.125, 0.0],
[0.9, 0.5, -0.25, -0.125, 0.0]],
[[0.9, 0.5, -0.25, -1.06, 0.0], [0.9, 0.5, -0.25, -1.12, 0.0]],
],
dtype=np.float32,
)
sample_rate = 16000
augment = PeakNormalization(apply_to="all", p=1.0)
processed_samples = augment(samples=torch.from_numpy(samples),
sample_rate=sample_rate).numpy()
assert_almost_equal(
processed_samples,
np.array(
[
np.array([[0.75, 0.5, -0.25, -0.125, 0.0],
[0.65, 0.5, -0.25, -0.125, 0.0]]) / 0.75,
np.array([[0.3, 0.5, -0.25, -0.125, 0.0],
[0.9, 0.5, -0.25, -0.125, 0.0]]) / 0.9,
np.array([[0.9, 0.5, -0.25, -1.06, 0.0],
[0.9, 0.5, -0.25, -1.12, 0.0]]) / 1.12,
],
dtype=np.float32,
),
)
self.assertEqual(processed_samples.dtype, np.float32)
def test_freeze_parameters(self):
samples1 = np.array(
[[[0.9, 0.5, -0.25, -0.125, 0.0]], [[0.9, 0.5, -0.25, -1.12, 0.0]]],
dtype=np.float32,
)
samples2 = np.array(
[[[0.1, -0.2, -0.35, -0.625, 2.0]], [[0.2, 0.9, -0.05, -0.12, 0.0]]],
dtype=np.float32,
)
sample_rate = 16000
augment = PeakNormalization(p=1.0, output_type="dict")
_ = augment(
samples=torch.from_numpy(samples1), sample_rate=sample_rate
).samples.numpy()
augment.freeze_parameters()
processed_samples2 = augment(
samples=torch.from_numpy(samples2), sample_rate=sample_rate
).samples.numpy()
augment.unfreeze_parameters()
assert_almost_equal(
processed_samples2,
np.array(
[
[[0.1 / 0.9, -0.2 / 0.9, -0.35 / 0.9, -0.625 / 0.9, 2.0 / 0.9]],
[[0.2 / 1.12, 0.9 / 1.12, -0.05 / 1.12, -0.12 / 1.12, 0.0 / 1.12]],
],
dtype=np.float32,
),
)
self.assertEqual(processed_samples2.dtype, np.float32)
def test_only_digital_silence(self):
"""Check that an exception is not thrown is selector is all False."""
samples = np.array(
[[0.0, 0.0, 0.0, 0.0, 0.0], [0.0, 0.0, 0.0, 0.0, 0.0]], dtype=np.float32
)
sample_rate = 16000
augment = PeakNormalization(p=1.0)
processed_samples = augment(
samples=torch.from_numpy(samples), sample_rate=sample_rate
).numpy()
assert_almost_equal(
processed_samples,
np.array(
[[0.0, 0.0, 0.0, 0.0, 0.0], [0.0, 0.0, 0.0, 0.0, 0.0]], dtype=np.float32
),
)
self.assertEqual(processed_samples.dtype, np.float32)
def test_variability_within_batch(self):
samples = np.array([[0.75, 0.5, 0.25, 0.125, 0.01]], dtype=np.float32)
samples_batch = np.stack([samples] * 1337, axis=0)
sample_rate = 16000
augment = PeakNormalization(p=0.5, output_type="dict")
processed_samples = augment(
samples=torch.from_numpy(samples_batch), sample_rate=sample_rate
).samples.numpy()
self.assertEqual(processed_samples.dtype, np.float32)
num_unprocessed_examples = 0
num_processed_examples = 0
for i in range(processed_samples.shape[0]):
if np.allclose(processed_samples[i], samples_batch[i]):
num_unprocessed_examples += 1
else:
num_processed_examples += 1
self.assertEqual(num_unprocessed_examples + num_processed_examples, 1337)
self.assertGreater(num_processed_examples, 0.2 * 1337)
self.assertLess(num_processed_examples, 0.8 * 1337)