def test_keras_model_preprocess():
num_classes = 1000
bounds = (0, 255)
channels = num_classes
with warnings.catch_warnings():
warnings.filterwarnings("ignore", category=DeprecationWarning)
inputs = Input(shape=(5, 5, channels))
logits = GlobalAveragePooling2D(data_format="channels_last")(inputs)
preprocessing = (
np.arange(num_classes)[None, None],
np.random.uniform(size=(5, 5, channels)) + 1,
)
model1 = KerasModel(Model(inputs=inputs, outputs=logits),
bounds=bounds,
predicts="logits")
model2 = KerasModel(
Model(inputs=inputs, outputs=logits),
bounds=bounds,
predicts="logits",
preprocessing=preprocessing,
)
model3 = KerasModel(Model(inputs=inputs, outputs=logits),
bounds=bounds,
predicts="logits")
preprocessing = (0, np.random.uniform(size=(5, 5, channels)) + 1)
model4 = KerasModel(
Model(inputs=inputs, outputs=logits),
bounds=bounds,
predicts="logits",
preprocessing=preprocessing,
)
np.random.seed(22)
test_images = np.random.rand(2, 5, 5, channels).astype(np.float32)
test_images_copy = test_images.copy()
p1 = model1.forward(test_images)
p2 = model2.forward(test_images)
# make sure the images have not been changed by
# the in-place preprocessing
assert np.all(test_images == test_images_copy)
p3 = model3.forward(test_images)
assert p1.shape == p2.shape == p3.shape == (2, num_classes)
np.testing.assert_array_almost_equal(p1 - p1.max(),
p3 - p3.max(),
decimal=5)
model4.forward(test_images)
def test_keras_model_probs(num_classes):
bounds = (0, 255)
channels = num_classes
with warnings.catch_warnings():
warnings.filterwarnings("ignore", category=DeprecationWarning)
inputs = Input(shape=(5, 5, channels))
logits = GlobalAveragePooling2D(data_format="channels_last")(inputs)
probs = Activation(softmax)(logits)
model1 = KerasModel(Model(inputs=inputs, outputs=logits),
bounds=bounds,
predicts="logits")
model2 = KerasModel(Model(inputs=inputs, outputs=probs),
bounds=bounds,
predicts="probabilities")
model3 = KerasModel(Model(inputs=inputs, outputs=probs),
bounds=bounds,
predicts="probs")
np.random.seed(22)
test_images = np.random.rand(2, 5, 5, channels).astype(np.float32)
p1 = model1.forward(test_images)
p2 = model2.forward(test_images)
p3 = model3.forward(test_images)
assert p1.shape == p2.shape == p3.shape == (2, num_classes)
np.testing.assert_array_almost_equal(p1 - p1.max(),
p2 - p2.max(),
decimal=1)
np.testing.assert_array_almost_equal(p2 - p2.max(),
p3 - p3.max(),
decimal=5)
