def test_serialize_deserialize_transform_speed(audio):
speed_orig = Speed(factor=1.1)
data_speed = speed_orig.to_dict()
speed = AudioTransform.from_dict(data_speed)
perturbed_speed = speed(audio, SAMPLING_RATE)
assert perturbed_speed.shape == (1, 14545)
def test_deserialize_transform(audio):
speed = AudioTransform.from_dict({
'name': 'Speed',
'kwargs': {
'factor': 1.1
}
})
perturbed = speed(audio, SAMPLING_RATE)
assert perturbed.shape == (1, 14545)
def test_speed_does_not_change_num_samples(audio):
augment_fn = SoxEffectTransform(effects=speed(SAMPLING_RATE))
# Since speed() is not deterministic and between 0.9x - 1.1x, multiple invocations
# will yield either slower (more samples) or faster (less samples) signal.
# The truncation/padding is performed inside of SoxEffectTransform so the user should not
# see these changes.
for _ in range(10):
augmented_audio = augment_fn(audio, sampling_rate=SAMPLING_RATE)
assert augmented_audio.shape == audio.shape
assert augmented_audio != audio
def test_deserialize_transform_speed(audio):
speed = AudioTransform.from_dict({
"name": "Speed",
"kwargs": {
"factor": 1.1
}
})
perturbed_speed = speed(audio, SAMPLING_RATE)
assert perturbed_speed.shape == (1, 14545)
def test_resample(audio, sampling_rate):
speed = Resample(source_sampling_rate=16000, target_sampling_rate=sampling_rate)
perturbed = speed(audio, SAMPLING_RATE)
assert perturbed.shape == (1, sampling_rate)
def test_speed(audio):
speed = Speed(factor=1.1)
perturbed = speed(audio, SAMPLING_RATE)
assert perturbed.shape == (1, 14545)
