def __getitem__(self, idx):
i = idx * batch_size
length = min(batch_size, (len(self.samples) - i))
batch_inputs = np.empty((length, img_height, img_width, 3),
dtype=np.float32)
batch_target = np.empty((length, num_classes), dtype=np.float32)
for i_batch in range(length):
sample = self.samples[i + i_batch]
image_id = sample['image_id']
filename = os.path.join(self.image_folder, image_id)
class_id = sample['disease_class']
image = cv.imread(filename) # BGR
# print(filename)
# print(image.shape)
image = cv.resize(image, (img_width, img_height), cv.INTER_CUBIC)
image = image[:, :, ::-1] # RGB
if self.usage == 'train':
image = aug_pipe.augment_image(image)
batch_inputs[i_batch] = preprocess_input(image)
batch_target[i_batch] = to_categorical(class_id, num_classes)
return batch_inputs, batch_target
def __getitem__(self, idx):
i = idx * batch_size
length = min(batch_size, (len(self.samples) - i))
batch_inputs = np.empty((3, length, img_size, img_size, channel), dtype=np.float32)
batch_dummy_target = np.zeros((length, embedding_size * 3), dtype=np.float32)
for i_batch in range(length):
sample = self.samples[i + i_batch]
for j, role in enumerate(['a', 'p', 'n']):
image_name = sample[role]
filename = os.path.join(self.image_folder, image_name)
image = cv.imread(filename) # BGR
image = image[:, :, ::-1] # RGB
dets = self.detector(image, 1)
num_faces = len(dets)
if num_faces > 0:
# Find the 5 face landmarks we need to do the alignment.
faces = dlib.full_object_detections()
for detection in dets:
faces.append(self.sp(image, detection))
image = dlib.get_face_chip(image, faces[0], size=img_size)
else:
image = cv.resize(image, (img_size, img_size), cv.INTER_CUBIC)
if self.usage == 'train':
image = aug_pipe.augment_image(image)
batch_inputs[j, i_batch] = preprocess_input(image)
return [batch_inputs[0], batch_inputs[1], batch_inputs[2]], batch_dummy_target
def __getitem__(self, idx):
i = idx * batch_size
length = min(batch_size, (len(self.samples) - i))
batch_inputs = np.empty((3, length, img_size, img_size, channel), dtype=np.float32)
batch_dummy_target = np.zeros((length, embedding_size * 3), dtype=np.float32)
for i_batch in range(length):
sample = self.samples[i + i_batch]
for j, role in enumerate(['a', 'p', 'n']):
image_name = sample[role]
filename = os.path.join(self.image_folder, self.usage, image_name)
image = cv.imread(filename) # BGR
image = image[:, :, ::-1] # RGB
image = cv.resize(image, (img_size, img_size), cv.INTER_CUBIC)
if self.usage == 'train':
image = aug_pipe.augment_image(image)
batch_inputs[j, i_batch] = preprocess_input(image)
return [batch_inputs[0], batch_inputs[1], batch_inputs[2]], batch_dummy_target