def classify(model, test_dir):
"""
Classifies all images in test_dir
:param model: Model to be evaluated
:param test_dir: Directory including the images
:return: None
"""
test_img_gen = image_generator.flow_from_directory(
test_dir,
target_size=(IMG_SIZE, IMG_SIZE),
classes=LABELS,
batch_size=1,
shuffle=False,
)
######### Your code starts here #########
# Classify all images in the given folder
# Calculate the accuracy and the number of test samples in the folder
# test_img_gen has a list attribute filenames where you can access the
# filename of the datapoint
# break after 1 epoch
#pdb.set_trace()
######### Your code ends here #########
print(f"Evaluated on {num_test} samples.")
print(f"Accuracy: {accuracy*100:.0f}%")
def get_bottleneck_dataset(model, img_dir, img_size):
# image_generator is of type ImageDataGenerator
# Find more information here
# https://www.tensorflow.org/api_docs/python/tf/keras/preprocessing/image/ImageDataGenerator?version=stable
train_img_gen = image_generator.flow_from_directory(
img_dir,
target_size=(img_size, img_size),
shuffle=False, # this is important for accessing filenames
classes=LABELS,
batch_size=1)
bottelneck_x_l = []
bottelneck_y_l = []
for i in range(train_img_gen.samples):
######### Your code starts here #########
# We want to create a dataset of bottleneck data.
# You can get the next image input and one-hot encoded label via x_i, y_i = next(train_img_gen)
# For each iteration append the bottleneck output as well as the label to the respective list
# bottelneck_x_l -> list of tensors with dimension [1, bottleneck_size]
# bottelneck_y_l -> list of tensors with dimension [1, num_labels]
# Fill in the parts indicated by #FILL#. No additional lines are required.
x_i, y_i = next(train_img_gen)
bottelneck_x = model(x_i)
bottelneck_x_l.append(bottelneck_x)
bottelneck_y_l.append(y_i)
######### Your code ends here #########
bottelneck_ds = tf.data.Dataset.from_tensor_slices(
(np.vstack(bottelneck_x_l), np.vstack(bottelneck_y_l)))
return bottelneck_ds, train_img_gen.samples
def classify(model, test_dir):
"""
Classifies all images in test_dir
:param model: Model to be evaluated
:param test_dir: Directory including the images
:return: None
"""
test_img_gen = image_generator.flow_from_directory(test_dir,
target_size=(IMG_SIZE,
IMG_SIZE),
classes=LABELS,
batch_size=1,
shuffle=False)
######### Your code starts here #########
# Classify all images in the given folder
# Calculate the accuracy and the number of test samples in the folder
# test_img_gen has a list attribute filenames where you can access the filename of the datapoint
num_test = test_img_gen.samples
correct_count = 0
for i in range(num_test):
x_i, y_i = next(test_img_gen)
y_pred = model(x_i).numpy()
if np.argmax(y_pred[0]) == np.argmax(y_i[0]):
correct_count += 1
else:
print(test_img_gen.filenames[i])
######### Your code ends here #########
accuracy = correct_count / num_test
print(f"Evaluated on {num_test} samples.")
print(f"Accuracy: {accuracy*100:.0f}%")
def get_bottleneck_dataset(model, img_dir, img_size):
# image_generator is of type ImageDataGenerator
# Find more information here
# https://www.tensorflow.org/api_docs/python/tf/keras/preprocessing/image/ImageDataGenerator?version=stable
train_img_gen = image_generator.flow_from_directory(img_dir,
target_size=(img_size, img_size),
shuffle=False, # this is important for accessing filenames
classes=LABELS,
batch_size=1)
bottelneck_x_l = []
bottelneck_y_l = []
for i in range(train_img_gen.samples):
######### Your code starts here #########
# We want to create a dataset of bottleneck data.
# You can get the next image input and one-hot encoded label via x_i, y_i = next(train_img_gen)
# For each iteration append the bottleneck output as well as the label to the respective list
# bottelneck_x_l -> list of tensors with dimension [1, bottleneck_size]
# bottelneck_y_l -> list of tensors with dimension [1, num_labels]
# Fill in the parts indicated by #FILL#. No additional lines are required.
######### Your code ends here #########
bottelneck_ds = tf.data.Dataset.from_tensor_slices((np.vstack(bottelneck_x_l),
np.vstack(bottelneck_y_l)))
return bottelneck_ds, train_img_gen.samples
def retrain(image_dir):
# Create the base model from the pre-trained model InceptionV3
base_model = tf.keras.applications.InceptionV3(input_shape=IMG_SHAPE,
include_top=False,
pooling='avg',
weights='imagenet')
base_model.summary()
base_model.compile(loss='mse')
print("Generating Bottleneck Dataset... this may take some minutes.")
bottleneck_train_ds, num_train = get_bottleneck_dataset(base_model, img_dir=image_dir, img_size=IMG_SIZE)
train_batches = bottleneck_train_ds.shuffle(10000).batch(BATCH_SIZE).repeat()
print("Done generating Bottleneck Dataset")
######### Your code starts here #########
# We want to create a linear classifier which takes the bottleneck data as input
# 1. Get the size of the bottleneck tensor. Hint: You can get the shape of a tensor via tensor.get_shape().as_list()
# 2. Define a new tf.keras Model which is a linear classifier
# 2.1 Define a keras Input (retrain_input)
# 2.2 Define the trainable layer (retrain_layer)
# 2.3 Define the activation function (retrain activation)
# 2.4 Create a new model
# 3. Define a loss and a evaluation metric
# Fill in the parts indicated by #FILL#. No additional lines are required.
######### Your code ends here #########
retrain_model.compile(optimizer=tf.keras.optimizers.SGD(lr=lr),
loss=loss,
metrics=[metric])
retrain_model.summary()
EPOCHS = 1
steps_per_epoch = 5000
tensorboard_callback = tf.keras.callbacks.TensorBoard(log_dir='./retrain_logs', update_freq='batch')
retrain_model.fit(train_batches,
epochs=EPOCHS,
steps_per_epoch=steps_per_epoch,
callbacks=[tensorboard_callback])
######### Your code starts here #########
# We now want to create the full model using the newly trained classifier
# Use tensorflow keras Sequential to stack the base_model and the new layers
# Fill in the parts indicated by #FILL#. No additional lines are required.
######### Your code ends here #########
model.compile(loss=loss, metrics=[metric])
maybe_makedirs('./trained_models')
model.save('./trained_models/trained.h5')
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--image_dir', type=str)
FLAGS, _ = parser.parse_known_args()
retrain(FLAGS.image_dir)