-
Notifications
You must be signed in to change notification settings - Fork 320
/
augment.py
506 lines (410 loc) · 18.2 KB
/
augment.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
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
# Copyright (c) 2021 PPViT Authors. All Rights Reserved.
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""Augmentation
RandAug:
- reference: RandAugment: Practical automated data augmentation with a reduced search space
- https://arxiv.org/abs/1909.13719
AutoAug:
- reference: AutoAugment: Learning Augmentation Policies from Data
- https://arxiv.org/abs/1805.09501
"""
import random
import numpy as np
from PIL import Image, ImageEnhance, ImageOps
LEVEL_DENOM = 10
#fill color is set to 128 instead fo image mean
def auto_augment_policy_v0():
"""policy v0: hack from timm"""
# ImageNet v0 policy from TPU EfficientNet impl, cannot find a paper reference.
policy = [
[('Equalize', 0.8, 1), ('ShearY', 0.8, 4)],
[('Color', 0.4, 9), ('Equalize', 0.6, 3)],
[('Color', 0.4, 1), ('Rotate', 0.6, 8)],
[('Solarize', 0.8, 3), ('Equalize', 0.4, 7)],
[('Solarize', 0.4, 2), ('Solarize', 0.6, 2)],
[('Color', 0.2, 0), ('Equalize', 0.8, 8)],
[('Equalize', 0.4, 8), ('SolarizeAdd', 0.8, 3)],
[('ShearX', 0.2, 9), ('Rotate', 0.6, 8)],
[('Color', 0.6, 1), ('Equalize', 1.0, 2)],
[('Invert', 0.4, 9), ('Rotate', 0.6, 0)],
[('Equalize', 1.0, 9), ('ShearY', 0.6, 3)],
[('Color', 0.4, 7), ('Equalize', 0.6, 0)],
[('Posterize', 0.4, 6), ('AutoContrast', 0.4, 7)],
[('Solarize', 0.6, 8), ('Color', 0.6, 9)],
[('Solarize', 0.2, 4), ('Rotate', 0.8, 9)],
[('Rotate', 1.0, 7), ('TranslateYRel', 0.8, 9)],
[('ShearX', 0.0, 0), ('Solarize', 0.8, 4)],
[('ShearY', 0.8, 0), ('Color', 0.6, 4)],
[('Color', 1.0, 0), ('Rotate', 0.6, 2)],
[('Equalize', 0.8, 4), ('Equalize', 0.0, 8)],
[('Equalize', 1.0, 4), ('AutoContrast', 0.6, 2)],
[('ShearY', 0.4, 7), ('SolarizeAdd', 0.6, 7)],
[('Posterize', 0.8, 2), ('Solarize', 0.6, 10)],
[('Solarize', 0.6, 8), ('Equalize', 0.6, 1)],
[('Color', 0.8, 6), ('Rotate', 0.4, 5)],
]
policy = [[SubPolicy(*args) for args in subpolicy] for subpolicy in policy]
return policy
def auto_augment_policy_v0r():
"""policy v0r: hack from timm"""
# ImageNet v0 policy from TPU EfficientNet impl, with variation of Posterize used
# in Google research implementation (number of bits discarded increases with magnitude)
policy = [
[('Equalize', 0.8, 1), ('ShearY', 0.8, 4)],
[('Color', 0.4, 9), ('Equalize', 0.6, 3)],
[('Color', 0.4, 1), ('Rotate', 0.6, 8)],
[('Solarize', 0.8, 3), ('Equalize', 0.4, 7)],
[('Solarize', 0.4, 2), ('Solarize', 0.6, 2)],
[('Color', 0.2, 0), ('Equalize', 0.8, 8)],
[('Equalize', 0.4, 8), ('SolarizeAdd', 0.8, 3)],
[('ShearX', 0.2, 9), ('Rotate', 0.6, 8)],
[('Color', 0.6, 1), ('Equalize', 1.0, 2)],
[('Invert', 0.4, 9), ('Rotate', 0.6, 0)],
[('Equalize', 1.0, 9), ('ShearY', 0.6, 3)],
[('Color', 0.4, 7), ('Equalize', 0.6, 0)],
[('PosterizeIncreasing', 0.4, 6), ('AutoContrast', 0.4, 7)],
[('Solarize', 0.6, 8), ('Color', 0.6, 9)],
[('Solarize', 0.2, 4), ('Rotate', 0.8, 9)],
[('Rotate', 1.0, 7), ('TranslateYRel', 0.8, 9)],
[('ShearX', 0.0, 0), ('Solarize', 0.8, 4)],
[('ShearY', 0.8, 0), ('Color', 0.6, 4)],
[('Color', 1.0, 0), ('Rotate', 0.6, 2)],
[('Equalize', 0.8, 4), ('Equalize', 0.0, 8)],
[('Equalize', 1.0, 4), ('AutoContrast', 0.6, 2)],
[('ShearY', 0.4, 7), ('SolarizeAdd', 0.6, 7)],
[('PosterizeIncreasing', 0.8, 2), ('Solarize', 0.6, 10)],
[('Solarize', 0.6, 8), ('Equalize', 0.6, 1)],
[('Color', 0.8, 6), ('Rotate', 0.4, 5)],
]
policy = [[SubPolicy(*args) for args in subpolicy] for subpolicy in policy]
return policy
def auto_augment_policy_originalr():
"""policy originalr: hack from timm"""
# ImageNet policy from https://arxiv.org/abs/1805.09501 with research posterize variation
policy = [
[('PosterizeIncreasing', 0.4, 8), ('Rotate', 0.6, 9)],
[('Solarize', 0.6, 5), ('AutoContrast', 0.6, 5)],
[('Equalize', 0.8, 8), ('Equalize', 0.6, 3)],
[('PosterizeIncreasing', 0.6, 7), ('PosterizeIncreasing', 0.6, 6)],
[('Equalize', 0.4, 7), ('Solarize', 0.2, 4)],
[('Equalize', 0.4, 4), ('Rotate', 0.8, 8)],
[('Solarize', 0.6, 3), ('Equalize', 0.6, 7)],
[('PosterizeIncreasing', 0.8, 5), ('Equalize', 1.0, 2)],
[('Rotate', 0.2, 3), ('Solarize', 0.6, 8)],
[('Equalize', 0.6, 8), ('PosterizeIncreasing', 0.4, 6)],
[('Rotate', 0.8, 8), ('Color', 0.4, 0)],
[('Rotate', 0.4, 9), ('Equalize', 0.6, 2)],
[('Equalize', 0.0, 7), ('Equalize', 0.8, 8)],
[('Invert', 0.6, 4), ('Equalize', 1.0, 8)],
[('Color', 0.6, 4), ('Contrast', 1.0, 8)],
[('Rotate', 0.8, 8), ('Color', 1.0, 2)],
[('Color', 0.8, 8), ('Solarize', 0.8, 7)],
[('Sharpness', 0.4, 7), ('Invert', 0.6, 8)],
[('ShearX', 0.6, 5), ('Equalize', 1.0, 9)],
[('Color', 0.4, 0), ('Equalize', 0.6, 3)],
[('Equalize', 0.4, 7), ('Solarize', 0.2, 4)],
[('Solarize', 0.6, 5), ('AutoContrast', 0.6, 5)],
[('Invert', 0.6, 4), ('Equalize', 1.0, 8)],
[('Color', 0.6, 4), ('Contrast', 1.0, 8)],
[('Equalize', 0.8, 8), ('Equalize', 0.6, 3)],
]
policy = [[SubPolicy(*args) for args in subpolicy] for subpolicy in policy]
return policy
def auto_augment_policy_original():
"""25 types of augment policies in original paper"""
policy = [
[('PosterizeOriginal', 0.4, 8), ('Rotate', 0.6, 9)],
[('Solarize', 0.6, 5), ('AutoContrast', 0.6, 5)],
[('Equalize', 0.8, 8), ('Equalize', 0.6, 3)],
[('PosterizeOriginal', 0.6, 7), ('PosterizeOriginal', 0.6, 6)],
[('Equalize', 0.4, 7), ('Solarize', 0.2, 4)],
[('Equalize', 0.4, 4), ('Rotate', 0.8, 8)],
[('Solarize', 0.6, 3), ('Equalize', 0.6, 7)],
[('PosterizeOriginal', 0.8, 5), ('Equalize', 1.0, 2)],
[('Rotate', 0.2, 3), ('Solarize', 0.6, 8)],
[('Equalize', 0.6, 8), ('PosterizeOriginal', 0.4, 6)],
[('Rotate', 0.8, 8), ('Color', 0.4, 0)],
[('Rotate', 0.4, 9), ('Equalize', 0.6, 2)],
[('Equalize', 0.0, 7), ('Equalize', 0.8, 8)],
[('Invert', 0.6, 4), ('Equalize', 1.0, 8)],
[('Color', 0.6, 4), ('Contrast', 1.0, 8)],
[('Rotate', 0.8, 8), ('Color', 1.0, 2)],
[('Color', 0.8, 8), ('Solarize', 0.8, 7)],
[('Sharpness', 0.4, 7), ('Invert', 0.6, 8)],
[('ShearX', 0.6, 5), ('Equalize', 1.0, 9)],
[('Color', 0.4, 0), ('Equalize', 0.6, 3)],
[('Equalize', 0.4, 7), ('Solarize', 0.2, 4)],
[('Solarize', 0.6, 5), ('AutoContrast', 0.6, 5)],
[('Invert', 0.6, 4), ('Equalize', 1.0, 8)],
[('Color', 0.6, 4), ('Contrast', 1.0, 8)],
[('Equalize', 0.8, 8), ('Equalize', 0.6, 3)],
]
policy = [[SubPolicy(*args) for args in subpolicy] for subpolicy in policy]
return policy
class AutoAugment():
"""Auto Augment
Randomly choose a tuple of augment ops from a list of policy
Then apply the tuple of augment ops to input image
Examples:
policy = auto_augment_policy_original()
augment = AutoAugment(policy)
transformed_image = augment(image)
"""
def __init__(self, policy):
self.policy = policy
def __call__(self, image, policy_idx=None):
if policy_idx is None:
policy_idx = random.randint(0, len(self.policy) - 1)
sub_policy = self.policy[policy_idx]
for operation in sub_policy:
image = operation(image)
return image
def rand_augment_policy_increasing(prob=0.5, magnitude_idx=9, magnitude_std=0.5):
"""
Rand augment policy: default rand-m9-mstd0.5-inc1
"""
policy = [
('AutoContrast', prob, magnitude_idx, magnitude_std),
('Equalize', prob, magnitude_idx, magnitude_std),
('Invert', prob, magnitude_idx, magnitude_std),
('Rotate', prob, magnitude_idx, magnitude_std),
('PosterizeIncreasing', prob, magnitude_idx, magnitude_std),
('SolarizeIncreasing', prob, magnitude_idx, magnitude_std),
('SolarizeAdd', prob, magnitude_idx, magnitude_std),
('ColorIncreasing', prob, magnitude_idx, magnitude_std),
('ContrastIncreasing', prob, magnitude_idx, magnitude_std),
('BrightnessIncreasing', prob, magnitude_idx, magnitude_std),
('SharpnessIncreasing', prob, magnitude_idx, magnitude_std),
('ShearX', prob, magnitude_idx, magnitude_std),
('ShearY', prob, magnitude_idx, magnitude_std),
('TranslateX', prob, magnitude_idx, magnitude_std),
('TranslateY', prob, magnitude_idx, magnitude_std),
]
policy = [SubPolicy(*args) for args in policy]
return policy
class RandAugment():
"""Rand Augment
Randomly choose N augment ops from a list of K policies
Then apply the N ops to input image
Examples:
policy = rand_augment_policy_original(magnitude_idx)
augment = RandAugment(policy)
transformed_image = augment(image)
"""
def __init__(self, policy, num_layers=2):
"""
Args:
policy: list of SubPolicy
num_layers: int
"""
self.policy = policy
self.num_layers = num_layers
def __call__(self, image):
selected_idx = np.random.choice(len(self.policy), self.num_layers)
for policy_idx in selected_idx:
sub_policy = self.policy[policy_idx]
image = sub_policy(image)
return image
class SubPolicy:
"""Subpolicy
Read augment name and magnitude, apply augment with probability
Args:
op_name: str, augment operation name
prob: float, if prob > random prob, apply augment
magnitude: int, index of magnitude in preset magnitude ranges
magnitude_std: float, std of magnitude in preset magnitude ranges
"""
def __init__(self, op_name, prob, magnitude, magnitude_std=0.5):
image_ops = {
'ShearX': shear_x,
'ShearY': shear_y,
'TranslateX': translate_x_absolute,
'TranslateY': translate_y_absolute,
'TranslateXRel': translate_x_relative,
'TranslateYRel': translate_y_relative,
'Rotate': rotate,
'AutoContrast': auto_contrast,
'Invert': invert,
'Equalize': equalize,
'Solarize': solarize,
'SolarizeIncreasing': solarize,
'SolarizeAdd': solarize_add,
'Posterize': posterize,
'PosterizeIncreasing': posterize,
'PosterizeOriginal': posterize,
'Contrast': contrast,
'ContrastIncreasing': contrast,
'Color': color,
'ColorIncreasing': color,
'Brightness': brightness,
'BrightnessIncreasing': brightness,
'Sharpness': sharpness,
'SharpnessIncreasing': sharpness,
}
level_fn = {
'ShearX': shear_level_to_arg,
'ShearY': shear_level_to_arg,
'TranslateX': translate_absolute_level_to_arg,
'TranslateY': translate_absolute_level_to_arg,
'TranslateXRel': translate_relative_level_to_arg,
'TranslateYRel': translate_relative_level_to_arg,
'Rotate': rotate_level_to_arg,
'AutoContrast': None,
'Invert': None,
'Equalize': None,
'Solarize': solarize_level_to_arg,
'SolarizeIncreasing': solarize_increasing_level_to_arg,
'SolarizeAdd': solarize_add_level_to_arg,
'Posterize': posterize_level_to_arg,
'PosterizeIncreasing': posterize_increasing_level_to_arg,
'PosterizeOriginal': posterize_original_level_to_arg,
'Contrast': enhance_level_to_arg,
'ContrastIncreasing': enhance_increasing_level_to_arg,
'Color': enhance_level_to_arg,
'ColorIncreasing': enhance_increasing_level_to_arg,
'Brightness': enhance_level_to_arg,
'BrightnessIncreasing': enhance_increasing_level_to_arg,
'Sharpness': enhance_level_to_arg,
'SharpnessIncreasing': enhance_increasing_level_to_arg,
}
self.prob = prob
self.magnitude = magnitude
self.magnitude_std = magnitude_std
self.ops = image_ops[op_name]
self.level_fn = level_fn[op_name]
def __call__(self, image):
if self.prob < 1.0 and random.random() > self.prob:
return image
magnitude = self.magnitude
# hack from timm auto_augment.py
if self.magnitude_std > 0:
if self.magnitude_std == float('inf'):
magnitude = random.uniform(0, magnitude)
elif self.magnitude_std > 0:
magnitude = random.gauss(magnitude, self.magnitude_std)
upper_bound = LEVEL_DENOM
magnitude = max(0, min(magnitude, upper_bound))
level_args = self.level_fn(magnitude) if self.level_fn is not None else tuple()
image = self.ops(image, *level_args)
return image
#################################################################
# Convert level to Image op arguments
#################################################################
def randomly_negate(value):
"""negate the value with 0.5 prob"""
return -value if random.random() > 0.5 else value
def shear_level_to_arg(level):
# range [-0.3, 0.3]
level = (level / LEVEL_DENOM) * 0.3
level = randomly_negate(level)
return (level,)
def translate_absolute_level_to_arg(level):
# translate const = 100
level = (level / LEVEL_DENOM) * 100.
level = randomly_negate(level)
return (level,)
def translate_relative_level_to_arg(level):
# range [-0.45, 0.45]
level = (level / LEVEL_DENOM) * 0.45
level = randomly_negate(level)
return (level,)
def rotate_level_to_arg(level):
# range [-30, 30]
level = (level / LEVEL_DENOM) * 30.
level = randomly_negate(level)
return (level,)
def solarize_level_to_arg(level):
# range [0, 256]
# intensity/severity of augmentation decreases with level
return (int((level / LEVEL_DENOM) * 256),)
def solarize_increasing_level_to_arg(level):
# range [0, 256]
# intensity/severity of augmentation increases with level
return (256 - int((level / LEVEL_DENOM) * 256),)
def solarize_add_level_to_arg(level):
# range [0, 110]
return (int((level / LEVEL_DENOM) * 110),)
def posterize_level_to_arg(level):
# range [0, 4]
# intensity/severity of augmentation decreases with level
return (int((level / LEVEL_DENOM) * 4),)
def posterize_increasing_level_to_arg(level):
# range [4, 0]
# intensity/severity of augmentation increases with level
return (4 - int((level / LEVEL_DENOM) * 4),)
def posterize_original_level_to_arg(level):
# range [4, 8]
# intensity/severity of augmentation decreases with level
return (int((level / LEVEL_DENOM) * 4) + 4,)
# For Contrast, Color, Brightness, Sharpness
def enhance_level_to_arg(level):
# range [0.1, 1.9]
return ((level / LEVEL_DENOM) * 1.8 + 0.1,)
# For ContrastIncreasing, ColorIncreasing, BrightnessIncreasing, SharpnessIncreasing
def enhance_increasing_level_to_arg(level):
# range [0.1, 1.9]
level = (level / LEVEL_DENOM) * 0.9
level = max(0.1, 1.0 + randomly_negate(level))
return (level,)
#################################################################
# PIL Image transforms
# https://pillow.readthedocs.io/en/stable/reference/Image.html#PIL.Image.Image.transform
#################################################################
def shear_x(image, factor, fillcolor=(128, 128, 128)):
return image.transform(image.size, Image.AFFINE, (1, factor, 0, 0, 1, 0), fillcolor=fillcolor)
def shear_y(image, factor, fillcolor=(128, 128, 128)):
return image.transform(image.size, Image.AFFINE, (1, 0, 0, factor, 1, 0), fillcolor=fillcolor)
def translate_x_absolute(image, pixels, fillcolor=(128, 128, 128)):
return image.transform(image.size, Image.AFFINE, (1, 0, pixels, 0, 1, 0), fillcolor=fillcolor)
def translate_y_absolute(image, pixels, fillcolor=(128, 128, 128)):
return image.transform(image.size, Image.AFFINE, (1, 0, 0, 0, 1, pixels), fillcolor=fillcolor)
def translate_x_relative(image, pct, fillcolor=(128, 128, 128)):
pixels = pct * image.size[0]
return image.transform(image.size, Image.AFFINE, (1, 0, pixels, 0, 1, 0), fillcolor=fillcolor)
def translate_y_relative(image, pct, fillcolor=(128, 128, 128)):
pixels = pct * image.size[0]
return image.transform(image.size, Image.AFFINE, (1, 0, 0, 0, 1, pixels), fillcolor=fillcolor)
def rotate(image, degrees):
return image.rotate(degrees)
def auto_contrast(image, magnitude=None):
return ImageOps.autocontrast(image)
def invert(image, magnitude=None):
return ImageOps.invert(image)
def equalize(image, magnitude=None):
return ImageOps.equalize(image)
def solarize(image, thresh):
return ImageOps.solarize(image, thresh)
def solarize_add(image, add, thresh=128):
lut = []
for i in range(256):
if i < thresh:
lut.append(min(255, i + add))
else:
lut.append(i)
if image.mode in ("L", "RGB"):
if image.mode == "RGB" and len(lut) == 256:
lut = lut + lut + lut
return image.point(lut)
return image
def posterize(image, bits_to_keep):
if bits_to_keep >= 8:
return image
return ImageOps.posterize(image, bits_to_keep)
def contrast(image, factor):
return ImageEnhance.Contrast(image).enhance(factor)
def color(image, factor):
return ImageEnhance.Color(image).enhance(factor)
def brightness(image, factor):
return ImageEnhance.Brightness(image).enhance(factor)
def sharpness(image, factor):
return ImageEnhance.Sharpness(image).enhance(factor)