def _transform(data, mean, train=True, mean_flag=False):
img, label = data
img = img.copy()
size316 = (316, 316)
size = (224, 224)
img = transforms.scale(img, 316)
img = transforms.center_crop(img, size316)
# 学習のときだけ実行
if train:
img = transforms.random_rotate(img)
img = transforms.random_flip(img, x_random=True, y_random=True)
# imgのリサイズ
img = img.transpose(1, 2, 0)
img = resize(img, size)
# 画像から平均を引く
if mean_flag:
img -= mean
print("mean use")
img *= (1.0 / 255.0)
img = img.transpose(2, 0, 1)
return img, label
def __call__(self, in_data):
img, bbox, label = in_data
_, H, W = img.shape
# random brightness and contrast
img = random_distort(img)
# rotate image
# return a tuple whose elements are rotated image, param.
# k (int in param)represents the number of times the image is rotated by 90 degrees.
img, params = transforms.random_rotate(img, return_param=True)
# restore the new hight and width
_, t_H, t_W = img.shape
# rotate bbox based on renewed parameters
bbox = rotate_bbox(bbox, (H, W), params['k'])
img = self.faster_rcnn.prepare(img)
# prepares the image to match the size of the image to be input into the RCNN
_, o_H, o_W = img.shape
# resize the bounding box according to the image resize
bbox = transforms.resize_bbox(bbox, (t_H, t_W), (o_H, o_W))
# horizontally & vertical flip
# simutaneously flip horizontally and vertically of the image
img, params = transforms.random_flip(img,
x_random=True,
y_random=True,
return_param=True)
# flip the bounding box with respect to the parameter
bbox = transforms.flip_bbox(bbox, (o_H, o_W),
x_flip=params['x_flip'],
y_flip=params['y_flip'])
scale = o_H / t_H
return img, bbox, label, scale
def _transform2(data, mean, train=True, mean_flag=False):
img, label = data
img = img.copy()
size316 = (316, 316)
size = (224, 224)
img_o = transforms.scale(img, 316)
img_o = transforms.center_crop(img_o, size316)
# 学習のときだけ実行
if train:
img_o = transforms.random_flip(img_o, y_random=True)
img_o = transforms.random_rotate(img_o)
# img = random_erase(img)
img_o = transforms.resize(img_o, size)
# 画像から平均を引く
if mean_flag:
img_o -= mean
img_o *= (1.0 / 255.0)
r = random.randint(316, 1500)
img_st = transforms.scale(img, r)
img_st = transforms.center_crop(img_st, (224, 224))
# 画像から平均を引く
if mean_flag:
img_st -= mean
img_st *= (1.0 / 255.0)
return img_o, label, img_st
def test_random_rotate(self):
img = np.random.uniform(size=(3, 24, 24))
out, rotation = random_rotate(img, return_rotation=True)
expected = np.transpose(img, axes=(1, 2, 0))
expected = np.rot90(expected, rotation)
expected = np.transpose(expected, axes=(2, 0, 1))
np.testing.assert_equal(out, expected)
def test_random_rotate(self):
img = np.random.uniform(size=(3, 24, 24))
out, param = random_rotate(img, return_param=True)
k = param['k']
expected = np.transpose(img, axes=(1, 2, 0))
expected = np.rot90(expected, k)
expected = np.transpose(expected, axes=(2, 0, 1))
np.testing.assert_equal(out, expected)
def __call__(self, in_data):
img, label = in_data
# img = transforms.resize(img, self.size)
img = transforms.random_crop(img, self.size)
img = transforms.flip(img, y_flip=self.y_flip, x_flip=self.x_flip)
img = transforms.random_rotate(img)
######################## perimage normalize #############
ms_im = img.reshape((img.shape[0], -1))
mean = np.mean(ms_im, axis=1)[:, np.newaxis, np.newaxis]
std = np.std(ms_im, axis=1)[:, np.newaxis, np.newaxis]
img = (img - mean) / (std + 0.0000001)
####################################################
return img, label
def augment_data(image, resize_width, resize_height, use_random_x_flip,
use_random_y_flip, use_random_rotate, use_pca_lighting,
crop_edit, crop_width, crop_height):
image = transforms.random_flip(image, use_random_x_flip, use_random_y_flip)
if use_random_rotate:
image = transforms.random_rotate(image)
image = transforms.pca_lighting(image, sigma=use_pca_lighting)
if crop_edit == 'Center Crop':
image = transforms.center_crop(image, (crop_width, crop_height))
elif crop_edit == 'Random Crop':
image = transforms.random_crop(image, (crop_width, crop_height))
image = transforms.resize(image, (resize_width, resize_height))
return image
def __call__(self, in_data):
img, label = in_data
# img = transforms.resize(img, self.size)
img = transforms.random_crop(img, self.size)
img = transforms.random_rotate(img)
img = transforms.random_flip(img,
x_random=True,
y_random=True,
return_param=False)
# aug = RandomBrightnessContrast(p=0.5) # WIP
# # aug = GaussNoise(p=0.5) # WIP
# img = aug(image=img)['image']
######################## perimage normalize #############
ms_im = img.reshape((img.shape[0], -1))
mean = np.mean(ms_im, axis=1)[:, np.newaxis, np.newaxis]
std = np.std(ms_im, axis=1)[:, np.newaxis, np.newaxis]
img = (img - mean) / (std + 0.0000001)
####################################################
return img, label
def get_example(self, i):
cutmix_alpha = self.cutmix_alpha
class_num = self.class_num
h = 224
w = 224
img_1, label_1 = self.base[i]
while True:
img_2, label_2 = random.choice(self.base)
if label_1 != label_2:
break
img_1 = transforms.scale(img_1, 316)
img_2 = transforms.scale(img_2, 316)
img_1 = transforms.center_crop(img_1, (316, 316))
img_2 = transforms.center_crop(img_2, (316, 316))
img_1 = transforms.random_flip(img_1, x_random=True, y_random=True)
img_2 = transforms.random_flip(img_2, x_random=True, y_random=True)
img_1 = img_1.transpose(1, 2, 0)
img_2 = img_2.transpose(1, 2, 0)
#img_1 = random_rotate(img_1)
#img_2 = random_rotate(img_2)
img_1 = img_1.transpose(2, 0, 1)
img_2 = img_2.transpose(2, 0, 1)
img_1 = transforms.resize(img_1, (224, 224))
img_2 = transforms.resize(img_2, (224, 224))
img_1 = transforms.random_rotate(img_1)
img_2 = transforms.random_rotate(img_2)
#####################################################
# cutmix実装
#####################################################
# sample the bounding box coordinates
while True:
# the combination ratio λ between two data points is
# sampled from the beta distribution Beta(α, α)
l = np.random.beta(cutmix_alpha, cutmix_alpha)
rx = random.randint(0, w)
rw = w * math.sqrt(1-l)
ry = random.randint(0, h)
rh = h * math.sqrt(1-l)
if ((ry + round(rh)) < h)and((rx + round(rw)) < w):
break
# denotes a binary mask indicating where to drop out
# and fill in from two images
M_1 = np.zeros((h, w))
M_1[ry:(ry + round(rh)),rx:(rx + round(rw))] = 1
M = np.ones((h, w))
M = M - M_1
# Define the combining operation.
img = (img_1 * M + img_2 * M_1).astype(np.float32)
# eye関数は単位行列を生成する
eye = np.eye(class_num)
label = (eye[label_1] * l + eye[label_2] * (1 - l)).astype(np.float32)
# 画像255で割る（輝度値をすべて0～1の間に収める）
img = img.transpose(1, 2, 0)
# 画像から平均を引く
if self.mean_flag:
img -= self.mean
img = img / 255.0
img = img.transpose(2, 0, 1)
return img, label
def __call__(self, in_data):
img, bbox, label = in_data
_, H, W = img.shape
# random brightness and contrast
img = random_distort(img)
# rotate image
# return a tuple whose elements are rotated image, param.
# k (int in param)represents the number of times the image is rotated by 90 degrees.
img, params = transforms.random_rotate(img, return_param=True)
# restore the new hight and width
_, t_H, t_W = img.shape
# rotate bbox based on renewed parameters
bbox = rotate_bbox(bbox, (H, W), params['k'])
# Random expansion:This method randomly place the input image on
# a larger canvas. The size of the canvas is (rH,rW), r is a random ratio drawn from [1,max_ratio].
# The canvas is filled by a value fill except for the region where the original image is placed.
if np.random.randint(2):
fill_value = img.mean(axis=1).mean(axis=1).reshape(-1, 1, 1)
img, param = transforms.random_expand(img,
max_ratio=2,
fill=fill_value,
return_param=True)
bbox = transforms.translate_bbox(bbox,
y_offset=param['y_offset'],
x_offset=param['x_offset'])
# Random crop
# crops the image with bounding box constraints
img, param = random_crop_with_bbox_constraints(img,
bbox,
min_scale=0.5,
max_aspect_ratio=1.5,
return_param=True)
# this translates bounding boxes to fit within the cropped area of an image, bounding boxes whose centers are outside of the cropped area are removed.
bbox, param = transforms.crop_bbox(bbox,
y_slice=param['y_slice'],
x_slice=param['x_slice'],
allow_outside_center=False,
return_param=True)
#assigning new labels to the bounding boxes after cropping
label = label[param['index']]
# if the bounding boxes are all removed,
if bbox.shape[0] == 0:
img, bbox, label = in_data
# update the height and width of the image
_, t_H, t_W = img.shape
img = self.faster_rcnn.prepare(img)
# prepares the image to match the size of the image to be input into the RCNN
_, o_H, o_W = img.shape
# resize the bounding box according to the image resize
bbox = transforms.resize_bbox(bbox, (t_H, t_W), (o_H, o_W))
# horizontally & vertical flip
# simutaneously flip horizontally and vertically of the image
img, params = transforms.random_flip(img,
x_random=True,
y_random=True,
return_param=True)
# flip the bounding box with respect to the parameter
bbox = transforms.flip_bbox(bbox, (o_H, o_W),
x_flip=params['x_flip'],
y_flip=params['y_flip'])
scale = o_H / t_H
return img, bbox, label, scale
