def load_data_to_buffer(self, data_file_range):
if not self._load_pre_processd_data:
tmp_x, self._y_buffer = get_data_from_files(
data_file_range=data_file_range,
model_type=self._params['model_type'])
#if input is an image, then we need to convert the string to float
if self._params['model_type'] == 'cnn' or self._params[
'model_type'] == 'siam':
self._x_buffer = []
for x_image in tmp_x:
if self._params['model_type'] == 'siam':
self._x_buffer.append(
(np.transpose(string2image(x_image[0][0]),
axes=[2, 1, 0]).flatten(),
np.transpose(string2image(x_image[1][0]),
axes=[2, 1, 0]).flatten()))
else:
self._x_buffer.append(
np.transpose(string2image(x_image[0]),
axes=[2, 1, 0]).flatten())
else:
self._x_buffer = tmp_x
else:
self._x_buffer, self._y_buffer = self._get_pre_processd_data.load_data(
)
assert (len(self._x_buffer) == len(self._y_buffer))
self._buffer_size = len(self._x_buffer)
def get_data(operation):
if operation == 'test':
data_file_indices = network_params_fwd['test_file_indices']
elif operation =='train':
data_file_indices = network_params_fwd['train_file_indices']
data_x, data_y = get_data_from_files(data_file_range=data_file_indices, model_type='fwd')
return data_x, data_y
def get_data(operation, string_img_convert=True):
if operation == 'test':
data_file_indices = network_params_cmbnd['test_file_indices']
elif operation == 'train':
data_file_indices = network_params_cmbnd['train_file_indices']
tmp_x, data_y = get_data_from_files(data_file_range=data_file_indices,
model_type='cnn')
data_x = []
if string_img_convert:
for x_image in tmp_x:
data_x.append(string2image(x_image[0]).flatten())
else:
data_x = tmp_x
return np.asarray(data_x), np.asarray(data_y)
def get_data(operation, string_img_convert=True):
raise Exception("Fix this damn thing")
if operation == 'test':
data_file_indices = network_params_siam['test_file_indices']
elif operation =='train':
data_file_indices = network_params_siam['train_file_indices']
tmp_x, data_y = get_data_from_files(data_file_range=data_file_indices, model_type='siam')
data_x = []
if string_img_convert:
for x_image in tmp_x:
data_x.append((np.transpose(string2image(x_image[0][0]), axes=[2,1,0]).flatten(),
np.transpose(string2image(x_image[1][0]), axes=[2,1,0]).flatten()))
else:
data_x = tmp_x
return np.asarray(data_x), np.asarray(data_y)
def get_data(data_file_indices, model_type):
tmp_x, y_buffer = get_data_from_files(data_file_range=data_file_indices,
model_type=model_type)
#if input is an image, then we need to convert the string to float
if model_type == 'cnn' or model_type == 'siam':
x_buffer = []
for x_image in tmp_x:
if model_type == 'siam':
x_buffer.append((np.transpose(string2image(x_image[0][0]),
axes=[2, 1, 0]).flatten(),
np.transpose(string2image(x_image[1][0]),
axes=[2, 1, 0]).flatten()))
else:
x_buffer.append(
np.transpose(string2image(x_image[0]), axes=[2, 1,
0]).flatten())
else:
x_buffer = tmp_x
assert (len(x_buffer) == len(y_buffer))
return x_buffer, y_buffer
