def __init__(self, folder_image, folder_label):
self.folder_image = folder_image
self.folder_label = folder_label
self.batch_size = BATCH_SIZE
self.max_txt_length = max_txt_length
self.examples = []
self.cur_index = 0
self.load_data()
self.image_util = ImageUtil(image_height=image_height, image_width=image_width)
self.vocab = Vocabulary()
class Generator:
def __init__(self, folder_image, folder_label):
self.folder_image = folder_image
self.folder_label = folder_label
self.batch_size = BATCH_SIZE
self.max_txt_length = max_txt_length
self.examples = []
self.cur_index = 0
self.load_data()
self.image_util = ImageUtil(image_height=image_height, image_width=image_width)
self.vocab = Vocabulary()
def load_data(self):
with open(self.folder_label, 'r') as f:
for line in f.readlines():
if ';' in line:
image_file, txt = line.split(sep=';', maxsplit=1)
image_file = os.path.abspath(os.path.join(self.folder_image, image_file))
txt = txt.strip()
if os.path.isfile(image_file):
self.examples.append((txt, image_file))
def examples_generator(self):
random.shuffle(self.examples)
while True:
images, target, encode_hidden = [], [], []
for i in range(self.batch_size):
self.cur_index += 1
if self.cur_index >= len(self.examples):
self.cur_index = 0
txt, img_path = self.examples[self.cur_index]
images.append(self.image_util.load(img_path))
target.append(self.vocab.one_hot_encode(txt))
# print(self.vocab.text_to_labels(txt))
# print(self.vocab.labels_to_text(target[0]))
yield np.array(images), np.array(target)
def main():
sess = tf.Session()
image_util = ImageUtil()
dataloader = ImageDataLoader(IMAGES_DIR, image_util)
train_dataset = dataloader.get_dataset('train', NUM_TRAIN_IMAGES)
val_dataset = dataloader.get_dataset('val', NUM_VAL_IMAGES)
test_dataset = dataloader.get_dataset('test', NUM_TEST_IMAGES)
model = ImageColorization().model
model.compile(optimizer=tf.train.AdamOptimizer(),
loss={'colorization_output': 'mean_squared_error', 'classification_output': 'categorical_crossentropy'},
loss_weights={'colorization_output': 1., 'classification_output': 1./300})
print 'Starting training'
for i in range(NUM_EPOCHS):
model.fit(train_dataset['greyscale'], {'colorization_output': train_dataset['color'], 'classification_output': train_dataset['class']}, epochs=1)
ab_unscaled = model.predict(val_dataset['greyscale'])[0]
lab = np.concatenate([val_dataset['greyscale']*100, ab_unscaled*200-100], axis=3)
show_predicted_values(ab_unscaled, lab)
print 'Finished training'
results = model.predict(test_dataset['greyscale'])
class ImageUtilityTest(unittest.TestCase):
"""Test to see if all the operations can be performed on the given file path."""
def setUp(self):
module_logger.debug('method setUp was called')
self.img = ImageUtil()
module_logger.debug('method setUp completed successfully')
def testImageLocation(self):
"""Testing for correct imagefile name extraction from url """
module_logger.debug('method testImageLocation was called')
name = '/name.png'
## Possible outcomes
outcomes = ['name.png', 'name.png.png', 'name.png.jpg', \
'name.png.gif', 'name.jpg', 'name.jpg.png', 'name.jpg.jpg', 'name.jpg.gif', \
'name.gif', 'name.gif.png', 'name.gif.jpg', 'name.gif.gif']
## Generate a SET of different kind of names and then verify that
## they all give the expected output
chars='~!@#$%^&*()_+-=`[]{}\|;:"<>,./?\'0jpngif'
#chars='abcdefghi'
for i in range(4):
name = chars[ random.randrange(len(chars)) ] + name
## The name is of format /____name.jpg****
for char1 in chars:
for char2 in chars:
for char3 in chars:
for char4 in chars:
self.assertIn(\
self.img.getImageFileName(name + char1 + char2 + char3 + char4), outcomes)
name = 'name.png'
for i in range(4):
name = name + chars[ random.randrange(len(chars)) ]
name = '/' + name
## name is of format /****name.jpg____
for char1 in chars:
for char2 in chars:
for char3 in chars:
for char4 in chars:
self.assertIn(\
self.img.getImageFileName(char1 + char2 + char3 + char4 + name), outcomes)
module_logger.debug('method testImageLocation completed successfully')
def testImageName(self):
"""Testing for correct image extensions"""
module_logger.debug('method testImageName was called')
self.assertEqual( self.img.getImageType('*.JPG'), 'jpg', 'JPG test failed')
self.assertEqual( self.img.getImageType('*.PNG'), 'png', 'PNG test failed')
self.assertEqual( self.img.getImageType('*.GIF'), 'gif', 'GIF test failed')
self.assertEqual( self.img.getImageType('***'), 'unknown Image Type', 'UNKNOWN IMAGE TYPE test failed')
module_logger.debug('method testImageName completed successfully')
def testImageConversion(self):
"""Testing for format conversion to JPG"""
module_logger.debug('method testImageConversion was called')
## Change the directory to get correct images
currentDir = getcwd()
workingDir = path.dirname( path.abspath(__file__) )
chdir(workingDir)
## Convert different formats to jpg
self.img.convertImageToJpg("JPEG.jpeg", "JPEG.jpg")
self.img.convertImageToJpg("PNG.png","PNG.jpg")
self.img.convertImageToJpg("JPG.jpg", "JPG.jpg")
self.img.convertImageToJpg("GIF.gif", "GIF.jpg")
## Check if the converted files are there
self.assertTrue( path.isfile("JPEG.jpg") )
self.assertTrue( path.isfile("PNG.jpg") )
self.assertTrue( path.isfile("JPG.jpg") )
self.assertTrue( path.isfile("GIF.jpg") )
## Delete the converted files
remove('JPEG.jpg')
remove('PNG.jpg')
remove('GIF.jpg')
chdir(currentDir)
module_logger.debug('method testImageConversion completed successfully')
def runTest(self):
filePointer = open('Test-Results.txt', 'a')
suite = unittest.TestLoader().loadTestsFromTestCase(ImageUtilityTest)
result = unittest.TextTestRunner(verbosity=2, stream=filePointer).run(suite)
filePointer.close()
return result
def setUp(self):
module_logger.debug('method setUp was called')
self.img = ImageUtil()
module_logger.debug('method setUp completed successfully')
cv2.imshow('origin', img)
def visual_attention(result, attention_plot):
len_result = len(result)
for i in range(len_result):
show_origin_image()
temp_att = np.reshape(attention_plot[i], (height, width))
cv2.imshow(f'predict word: {result[i]}', temp_att)
cv2.waitKey()
cv2.destroyAllWindows()
if __name__ == '__main__':
image_util = ImageUtil(image_height=image_height, image_width=image_width)
img_tensor = image_util.load(args.image)
img_tensor = np.expand_dims(img_tensor, 0)
result = ''
hidden = tf.zeros((1, decode_units))
word_one_hot = np.zeros((1, vocab_size))
word_one_hot[0][1] = 1.
attention_plot = np.zeros((max_txt_length, height * width))
for i in range(max_txt_length):
predict, hidden, attention_weights = model(word_one_hot, hidden,
img_tensor)
predict_id = tf.argmax(predict, axis=-1)
