forked from petrosgk/Kaggle-Carvana-Image-Masking-Challenge
-
Notifications
You must be signed in to change notification settings - Fork 1
/
train.py
185 lines (154 loc) · 7.89 KB
/
train.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
import cv2
import numpy as np
import pandas as pd
from keras.callbacks import EarlyStopping, ReduceLROnPlateau, ModelCheckpoint, TensorBoard
from sklearn.model_selection import train_test_split
import os
import random
gpu_id = '0'
os.environ['CUDA_VISIBLE_DEVICES']=str(gpu_id)
import params
DATA_PATH='/media/Disk/yanpengxiang/dataset/carvana/'
input_size = params.input_size
orig_width = params.orig_width
orig_height = params.orig_height
epochs = params.max_epochs
batch_size = params.batch_size
model = params.model_factory()
#df_train = pd.read_csv(DATA_PATH + 'train_masks.csv')
#ids_train = df_train['img'].map(lambda s: s.split('.')[0])
ids_train = []
with open('input/train_id.txt') as f:
for line in f:
ids_train.append(line[:15])
ids_train_split, ids_valid_split = train_test_split(ids_train, test_size=0.1, random_state=42)
print('Training on {} samples'.format(len(ids_train_split)))
print('Validating on {} samples'.format(len(ids_valid_split)))
def randomHueSaturationValue(image, hue_shift_limit=(-180, 180),
sat_shift_limit=(-255, 255),
val_shift_limit=(-255, 255), u=0.5):
if np.random.random() < u:
image = cv2.cvtColor(image, cv2.COLOR_BGR2HSV)
h, s, v = cv2.split(image)
hue_shift = np.random.uniform(hue_shift_limit[0], hue_shift_limit[1])
h = cv2.add(h, hue_shift)
sat_shift = np.random.uniform(sat_shift_limit[0], sat_shift_limit[1])
s = cv2.add(s, sat_shift)
val_shift = np.random.uniform(val_shift_limit[0], val_shift_limit[1])
v = cv2.add(v, val_shift)
image = cv2.merge((h, s, v))
image = cv2.cvtColor(image, cv2.COLOR_HSV2BGR)
return image
def randomShiftScaleRotate(image, mask,
shift_limit=(-0.0625, 0.0625),
scale_limit=(-0.1, 0.1),
rotate_limit=(-45, 45), aspect_limit=(0, 0),
borderMode=cv2.BORDER_CONSTANT, u=0.5):
if np.random.random() < u:
height, width, channel = image.shape
angle = np.random.uniform(rotate_limit[0], rotate_limit[1]) # degree
scale = np.random.uniform(1 + scale_limit[0], 1 + scale_limit[1])
aspect = np.random.uniform(1 + aspect_limit[0], 1 + aspect_limit[1])
sx = scale * aspect / (aspect ** 0.5)
sy = scale / (aspect ** 0.5)
dx = round(np.random.uniform(shift_limit[0], shift_limit[1]) * width)
dy = round(np.random.uniform(shift_limit[0], shift_limit[1]) * height)
cc = np.math.cos(angle / 180 * np.math.pi) * sx
ss = np.math.sin(angle / 180 * np.math.pi) * sy
rotate_matrix = np.array([[cc, -ss], [ss, cc]])
box0 = np.array([[0, 0], [width, 0], [width, height], [0, height], ])
box1 = box0 - np.array([width / 2, height / 2])
box1 = np.dot(box1, rotate_matrix.T) + np.array([width / 2 + dx, height / 2 + dy])
box0 = box0.astype(np.float32)
box1 = box1.astype(np.float32)
mat = cv2.getPerspectiveTransform(box0, box1)
image = cv2.warpPerspective(image, mat, (width, height), flags=cv2.INTER_LINEAR, borderMode=borderMode,
borderValue=(
0, 0,
0,))
mask = cv2.warpPerspective(mask, mat, (width, height), flags=cv2.INTER_LINEAR, borderMode=borderMode,
borderValue=(
0, 0,
0,))
return image, mask
def randomHorizontalFlip(image, mask, u=0.5):
if np.random.random() < u:
image = cv2.flip(image, 1)
mask = cv2.flip(mask, 1)
return image, mask
def train_generator():
while True:
for start in range(0, len(ids_train_split), batch_size):
x_batch = []
y_batch = []
end = min(start + batch_size, len(ids_train_split))
ids_train_batch = ids_train_split[start:end]
for id in ids_train_batch:
#rand_height = random.randint(0, orig_height - input_size - 1)
#rand_height = 0
#rand_width = random.randint(0, orig_width - input_size - 1)
img = cv2.imread((DATA_PATH + 'train/{}.jpg').format(id))
#img = img[rand_height:rand_height+input_size, rand_width:rand_width+input_size]
img = cv2.resize(img, (input_size, input_size))
mask = cv2.imread((DATA_PATH + 'gt/{}_mask.png').format(id), cv2.IMREAD_GRAYSCALE)
mask = cv2.resize(mask, (input_size, input_size))
#mask = mask[rand_height:rand_height+input_size, rand_width:rand_width+input_size]
img = randomHueSaturationValue(img,
hue_shift_limit=(-50, 50),
sat_shift_limit=(-5, 5),
val_shift_limit=(-15, 15))
img, mask = randomShiftScaleRotate(img, mask,
shift_limit=(-0.0625, 0.0625),
scale_limit=(-0.1, 0.1),
rotate_limit=(-0, 0))
img, mask = randomHorizontalFlip(img, mask)
mask = np.expand_dims(mask, axis=2)
x_batch.append(img)
y_batch.append(mask)
x_batch = np.array(x_batch, np.float32) / 255
y_batch = np.array(y_batch, np.float32) / 255
yield x_batch, y_batch
def valid_generator():
while True:
for start in range(0, len(ids_valid_split), batch_size):
x_batch = []
y_batch = []
end = min(start + batch_size, len(ids_valid_split))
ids_valid_batch = ids_valid_split[start:end]
for id in ids_valid_batch:
#rand_height = random.randint(0, orig_height - input_size - 1)
#rand_width = random.randint(0, orig_width - input_size - 1)
img = cv2.imread((DATA_PATH + 'train/{}.jpg').format(id))
img = cv2.resize(img, (input_size, input_size))
mask = cv2.imread((DATA_PATH + 'gt/{}_mask.png').format(id), cv2.IMREAD_GRAYSCALE)
#img = img[rand_height:rand_height+input_size, rand_width:rand_width+input_size]
mask = cv2.resize(mask, (input_size, input_size))
#mask = mask[rand_height:rand_height+input_size, rand_width:rand_width+input_size]
mask = np.expand_dims(mask, axis=2)
x_batch.append(img)
y_batch.append(mask)
x_batch = np.array(x_batch, np.float32) / 255
y_batch = np.array(y_batch, np.float32) / 255
yield x_batch, y_batch
callbacks = [EarlyStopping(monitor='val_loss',
patience=8,
verbose=1,
min_delta=1e-4),
ReduceLROnPlateau(monitor='val_loss',
factor=0.1,
patience=4,
verbose=1,
epsilon=1e-4),
ModelCheckpoint(monitor='val_loss',
filepath='weights/best_weights.hdf5',
save_best_only=True,
save_weights_only=True),
TensorBoard(log_dir='logs')]
model.load_weights(filepath='weights/best_weights.hdf5')
model.fit_generator(generator=train_generator(),
steps_per_epoch=np.ceil(float(len(ids_train_split)) / float(batch_size)),
epochs=epochs,
verbose=2,
callbacks=callbacks,
validation_data=valid_generator(),
validation_steps=np.ceil(float(len(ids_valid_split)) / float(batch_size)))