def test_restore_time_signal_with_str_window(self):
x = self.x
X = stft(x, window='hann')
tc.assert_almost_equal(
x, istft(X, 1024, 256, window='hann', num_samples=len(x)))
tc.assert_equal(X.shape, (154, 513))
def test_preemphasis(self):
y = self.x
yFilterd = transform.preemphasis(y, 0.97)
tc.assert_equal(yFilterd.shape, y.shape)
tc.assert_isreal(yFilterd)
tc.assert_equal(yFilterd[0], y[0])
def test_restore_time_signal_from_stft_and_istft(self):
x = self.time_signal
X = self.stft(self.torch_signal)
tc.assert_almost_equal(
self.stft.inverse(X)[..., :x.shape[-1]].numpy(), x)
tc.assert_equal(X.shape, (154, self.num_features))
def test_offcomp(self):
y = self.x
yFilterd = transform.offset_compensation(y)
tc.assert_equal(yFilterd.shape, y.shape)
tc.assert_isreal(yFilterd)
tc.assert_array_not_equal(yFilterd[1:], y[1:])
tc.assert_equal(yFilterd[0], y[0])
def test_mfcc(self):
path = get_file_path("sample.wav")
y = audioread(path)[0]
y_filtered = transform.mfcc(y)
tc.assert_equal(y_filtered.shape, (291, 13))
tc.assert_isreal(y_filtered)
def test_mfcc(self):
path = get_file_path("sample.wav")
y = audioread(path)[0]
yFilterd = transform.ssc(y)
tc.assert_equal(yFilterd.shape, (294, 26))
tc.assert_isreal(yFilterd)
def test_window_length(self):
X = stft(self.x, 512, 160, window_length=400)
x_hat = istft(X, 512, 160, window_length=400)
X_ref = istft(stft(self.x, 400, 160), 400, 160)
tc.assert_equal(X.shape, (243, 257))
tc.assert_allclose(X_ref, x_hat, rtol=1e-6, atol=1e-6)
def test_curated_input_with_keepdims_true(self):
labels = np.asarray([[0, 1, 2, 0]])
dtype = np.float32
expected_mask = np.asarray([[1, 0, 0, 1], [0, 1, 0, 0], [0, 0, 1, 0],
[0, 0, 0, 0], [0, 0, 0, 0]],
dtype=dtype)
tc.assert_equal(labels_to_one_hot(labels, categories=5, keepdims=True),
expected_mask)
def test_samples_to_stft_frames(self):
size = 1024
shift = 256
tc.assert_equal(_samples_to_stft_frames(1023, size, shift), 1)
tc.assert_equal(_samples_to_stft_frames(1024, size, shift), 1)
tc.assert_equal(_samples_to_stft_frames(1025, size, shift), 2)
tc.assert_equal(_samples_to_stft_frames(1024 + 256, size, shift), 2)
tc.assert_equal(_samples_to_stft_frames(1024 + 257, size, shift), 3)
def test_fbank(self):
path = get_file_path("sample.wav")
y = audioread(path)[0]
feature = transform.fbank(y)
tc.assert_equal(feature.shape, (240, 23))
tc.assert_isreal(feature)
tc.assert_array_greater_equal(feature, 0)
def test_dump_load(self, name, data, expect):
with tempfile.TemporaryDirectory() as temp_dir:
dump_hdf5(data, os.path.join(temp_dir, 'test.hdf5'))
data_load = load_hdf5(os.path.join(temp_dir, 'test.hdf5'))
assert 'key' in data_load.keys(), data_load
assert type(expect) is type(data_load['key']), \
(type(expect), type(data_load['key']), expect, data_load['key'])
tc.assert_equal(expect, data_load['key'])
def test_compare_both_biorthogonal_window_variants(self):
window = signal.blackman(1024)
shift = 256
for_result = _biorthogonal_window_loopy(window, shift)
vec_result = _biorthogonal_window(window, shift)
brute_force_result = _biorthogonal_window_brute_force(window, shift)
tc.assert_equal(for_result, vec_result)
tc.assert_allclose(for_result, brute_force_result)
tc.assert_equal(for_result.shape, (1024, ))
def test_cosine_similarity(self):
phi = 60 / 360 * 2 * np.pi
W1 = np.array([4, 0, 0])
W2 = np.array([1, 0, 0])
W3 = np.array([0, 1, 0])
W4 = np.array([1, 1j, 0])
W5 = np.array([1, -1j, 0])
W6 = np.array([np.cos(phi), np.sin(phi), 0])
tc.assert_equal(cos_similarity(W1, W1), 1.0)
tc.assert_equal(cos_similarity(W1, W2), 1.0)
tc.assert_equal(cos_similarity(W1, W3), 0.0)
tc.assert_equal(cos_similarity(W4, W5), 0.0)
tc.assert_equal(cos_similarity(W1, W6), np.cos(phi))
def test_spectrogram_and_energy(self):
x = self.x
X = stft(x)
spectrogram = stft_to_spectrogram(X)
energy = spectrogram_to_energy_per_frame(spectrogram)
tc.assert_equal(X.shape, (154, 513))
tc.assert_equal(spectrogram.shape, (154, 513))
tc.assert_isreal(spectrogram)
tc.assert_array_greater_equal(spectrogram, 0)
tc.assert_equal(energy.shape, (154, ))
tc.assert_isreal(energy)
tc.assert_array_greater_equal(energy, 0)
def test_stft_frame_count(self):
stft = self.stft
stft.fading = False
x = torch.rand(size=[1019])
X = stft(x)
tc.assert_equal(X.shape, (50, self.fbins * 2))
x = torch.rand(size=[1020])
X = stft(x)
tc.assert_equal(X.shape, (50, self.fbins * 2))
x = torch.rand(size=[1021])
X = stft(x)
tc.assert_equal(X.shape, (51, self.fbins * 2))
stft.fading = True
x = torch.rand(size=[1019])
X = stft(x)
tc.assert_equal(X.shape, (52, self.fbins * 2))
x = torch.rand(size=[1020])
X = stft(x)
tc.assert_equal(X.shape, (52, self.fbins * 2))
x = torch.rand(size=[1021])
X = stft(x)
tc.assert_equal(X.shape, (53, self.fbins * 2))
def test_uniform_scalar(self):
scalar = random.uniform()
tc.assert_equal(scalar.shape, ())
def test_hermitian_many_dimensions(self):
shape = (3, 4, 5, 2, 2)
matrix = random.hermitian(*shape)
tc.assert_equal(matrix.shape, shape)
tc.assert_equal(matrix, np.swapaxes(matrix, -1, -2).conj())
def test_hermitian_2D(self):
shape = (2, 2)
matrix = random.hermitian(*shape)
tc.assert_equal(matrix.shape, shape)
tc.assert_equal(matrix, np.swapaxes(matrix, -1, -2).conj())
def test_restore_time_signal_from_stft_and_istft(self):
x = self.x
X = stft(x)
tc.assert_almost_equal(x, istft(X, 1024, 256)[:len(x)])
tc.assert_equal(X.shape, (154, 513))
def test_stft_frames_to_samples(self):
size = 1024
shift = 256
tc.assert_equal(_stft_frames_to_samples(1, size, shift), 1024)
tc.assert_equal(_stft_frames_to_samples(2, size, shift), 1024 + 256)
def test_swap_and_reshape(self):
result = morph('T,B,F->T,F*B', A)
tc.assert_equal(result.shape, (T, F * B))
tc.assert_equal(result, A.swapaxes(-1, -2).reshape(T, F * B))
def test_transpose_and_reshape(self):
result = morph('T,B,F->F,B*T', A)
tc.assert_equal(result.shape, (F, B * T))
tc.assert_equal(result, A.transpose(2, 1, 0).reshape(F, B * T))
def test_reshape_and_broadcast_many(self):
result = morph('T,B,F->1,T,1,B*F,1', A)
tc.assert_equal(result.shape, (1, T, 1, B * F, 1))
def test_stft_frame_count(self):
stft = self.stft
stft.fading = False
x = torch.rand(size=[1023])
X = stft(x)
tc.assert_equal(X.shape, (1, self.num_features))
x = torch.rand(size=[1024])
X = stft(x)
tc.assert_equal(X.shape, (1, self.num_features))
x = torch.rand(size=[1025])
X = stft(x)
tc.assert_equal(X.shape, (2, self.num_features))
stft.fading = True
x = torch.rand(size=[1023])
X = stft(x)
tc.assert_equal(X.shape, (7, self.num_features))
x = torch.rand(size=[1024])
X = stft(x)
tc.assert_equal(X.shape, (7, self.num_features))
x = torch.rand(size=[1025])
X = stft(x)
tc.assert_equal(X.shape, (8, self.num_features))
def test_stft_frame_count(self):
stft_params = dict(size=1024, shift=256, fading=False)
x = np.random.normal(size=[1023])
X = stft(x, **stft_params)
tc.assert_equal(X.shape, (1, 513))
x = np.random.normal(size=[1024])
X = stft(x, **stft_params)
tc.assert_equal(X.shape, (1, 513))
x = np.random.normal(size=[1025])
X = stft(x, **stft_params)
tc.assert_equal(X.shape, (2, 513))
stft_params = dict(size=1024, shift=256, fading=True)
x = np.random.normal(size=[1023])
X = stft(x, **stft_params)
tc.assert_equal(X.shape, (7, 513))
x = np.random.normal(size=[1024])
X = stft(x, **stft_params)
tc.assert_equal(X.shape, (7, 513))
x = np.random.normal(size=[1025])
X = stft(x, **stft_params)
tc.assert_equal(X.shape, (8, 513))
stft_params = dict(size=512,
shift=160,
window_length=400,
fading=False)
x = np.random.normal(size=[399])
X = stft(x, **stft_params)
tc.assert_equal(X.shape, (1, 257))
x = np.random.normal(size=[400])
X = stft(x, **stft_params)
tc.assert_equal(X.shape, (1, 257))
x = np.random.normal(size=[401])
X = stft(x, **stft_params)
tc.assert_equal(X.shape, (2, 257))
x = np.random.normal(size=[559])
X = stft(x, **stft_params)
tc.assert_equal(X.shape, (2, 257))
x = np.random.normal(size=[560])
X = stft(x, **stft_params)
tc.assert_equal(X.shape, (2, 257))
x = np.random.normal(size=[561])
X = stft(x, **stft_params)
tc.assert_equal(X.shape, (3, 257))
def test_transpose_capital(self):
result = morph('tbB->tBb', A)
tc.assert_equal(result.shape, (T, F, B))
tc.assert_equal(result, A.transpose(0, 2, 1))
def test_restore_time_signal_from_stft_and_istft_kaldi_params(self):
x = self.x
X = stft(x, size=400, shift=160)
tc.assert_almost_equal(x, istft(X, 400, 160)[:len(x)])
tc.assert_equal(X.shape, (243, 201))
def test_restore_time_signal_from_stft_and_istft_with_num_samples(self):
x = self.x
X = stft(x)
tc.assert_almost_equal(x, istft(X, 1024, 256, num_samples=len(x)))
tc.assert_equal(X.shape, (154, 513))
def test_batch_mode(self):
size = 1024
shift = 256
# Reference
X = stft_single_channel(self.x)
x1 = np.array([self.x, self.x])
X1 = stft(x1)
tc.assert_equal(X1.shape, (2, 154, 513))
for d in np.ndindex(2):
tc.assert_equal(X1[d, :, :].squeeze(), X)
x11 = np.array([x1, x1])
X11 = stft(x11)
tc.assert_equal(X11.shape, (2, 2, 154, 513))
for d, k in np.ndindex(2, 2):
tc.assert_equal(X11[d, k, :, :].squeeze(), X)
x2 = x1.transpose()
X2 = stft(x2, axis=0)
tc.assert_equal(X2.shape, (154, 513, 2))
for d in np.ndindex(2):
tc.assert_equal(X2[:, :, d].squeeze(), X)
x21 = np.array([x2, x2])
X21 = stft(x21, axis=1)
tc.assert_equal(X21.shape, (2, 154, 513, 2))
for d, k in np.ndindex(2, 2):
tc.assert_equal(X21[d, :, :, k].squeeze(), X)
x22 = x21.swapaxes(0, 1)
X22 = stft(x22, axis=0)
tc.assert_equal(X22.shape, (154, 513, 2, 2))
for d, k in np.ndindex(2, 2):
tc.assert_equal(X22[:, :, d, k].squeeze(), X)
def test_reshape_and_broadcast(self):
tc.assert_equal(morph('T,B,F->T,1,B*F', A).shape, (T, 1, B * F))
tc.assert_equal(morph('T,B,F->T,1,B*F', A).ravel(), A.ravel())
