File size: 17,607 Bytes
28c6826
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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

""" MobileNet V3

A PyTorch impl of MobileNet-V3, compatible with TF weights from official impl.

Paper: Searching for MobileNetV3 - https://arxiv.org/abs/1905.02244

Hacked together by / Copyright 2020 Ross Wightman
"""
import torch
import torch.nn as nn
import torch.nn.functional as F

from typing import List

from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD, IMAGENET_INCEPTION_MEAN, IMAGENET_INCEPTION_STD
from .efficientnet_blocks import round_channels, resolve_bn_args, resolve_act_layer, BN_EPS_TF_DEFAULT
from .efficientnet_builder import EfficientNetBuilder, decode_arch_def, efficientnet_init_weights
from .features import FeatureInfo, FeatureHooks
from .helpers import build_model_with_cfg, default_cfg_for_features
from .layers import SelectAdaptivePool2d, Linear, create_conv2d, get_act_fn, hard_sigmoid
from .registry import register_model

__all__ = ['MobileNetV3']


def _cfg(url='', **kwargs):
    return {
        'url': url, 'num_classes': 1000, 'input_size': (3, 224, 224), 'pool_size': (1, 1),
        'crop_pct': 0.875, 'interpolation': 'bilinear',
        'mean': IMAGENET_DEFAULT_MEAN, 'std': IMAGENET_DEFAULT_STD,
        'first_conv': 'conv_stem', 'classifier': 'classifier',
        **kwargs
    }


default_cfgs = {
    'mobilenetv3_large_075': _cfg(url=''),
    'mobilenetv3_large_100': _cfg(
        interpolation='bicubic',
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/mobilenetv3_large_100_ra-f55367f5.pth'),
    'mobilenetv3_small_075': _cfg(url=''),
    'mobilenetv3_small_100': _cfg(url=''),
    'mobilenetv3_rw': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/mobilenetv3_100-35495452.pth',
        interpolation='bicubic'),
    'tf_mobilenetv3_large_075': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_mobilenetv3_large_075-150ee8b0.pth',
        mean=IMAGENET_INCEPTION_MEAN, std=IMAGENET_INCEPTION_STD),
    'tf_mobilenetv3_large_100': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_mobilenetv3_large_100-427764d5.pth',
        mean=IMAGENET_INCEPTION_MEAN, std=IMAGENET_INCEPTION_STD),
    'tf_mobilenetv3_large_minimal_100': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_mobilenetv3_large_minimal_100-8596ae28.pth',
        mean=IMAGENET_INCEPTION_MEAN, std=IMAGENET_INCEPTION_STD),
    'tf_mobilenetv3_small_075': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_mobilenetv3_small_075-da427f52.pth',
        mean=IMAGENET_INCEPTION_MEAN, std=IMAGENET_INCEPTION_STD),
    'tf_mobilenetv3_small_100': _cfg(
        url= 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_mobilenetv3_small_100-37f49e2b.pth',
        mean=IMAGENET_INCEPTION_MEAN, std=IMAGENET_INCEPTION_STD),
    'tf_mobilenetv3_small_minimal_100': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_mobilenetv3_small_minimal_100-922a7843.pth',
        mean=IMAGENET_INCEPTION_MEAN, std=IMAGENET_INCEPTION_STD),
}

_DEBUG = False


class MobileNetV3(nn.Module):
    """ MobiletNet-V3

    Based on my EfficientNet implementation and building blocks, this model utilizes the MobileNet-v3 specific
    'efficient head', where global pooling is done before the head convolution without a final batch-norm
    layer before the classifier.

    Paper: https://arxiv.org/abs/1905.02244
    """

    def __init__(self, block_args, num_classes=1000, in_chans=3, stem_size=16, num_features=1280, head_bias=True,
                 channel_multiplier=1.0, pad_type='', act_layer=nn.ReLU, drop_rate=0., drop_path_rate=0.,
                 se_kwargs=None, norm_layer=nn.BatchNorm2d, norm_kwargs=None, global_pool='avg'):
        super(MobileNetV3, self).__init__()

        self.num_classes = num_classes
        self.num_features = num_features
        self.drop_rate = drop_rate

        # Stem
        stem_size = round_channels(stem_size, channel_multiplier)
        self.conv_stem = create_conv2d(in_chans, stem_size, 3, stride=2, padding=pad_type)
        self.bn1 = norm_layer(stem_size, **norm_kwargs)
        self.act1 = act_layer(inplace=True)

        # Middle stages (IR/ER/DS Blocks)
        builder = EfficientNetBuilder(
            channel_multiplier, 8, None, 32, pad_type, act_layer, se_kwargs,
            norm_layer, norm_kwargs, drop_path_rate, verbose=_DEBUG)
        self.blocks = nn.Sequential(*builder(stem_size, block_args))
        self.feature_info = builder.features
        head_chs = builder.in_chs

        # Head + Pooling
        self.global_pool = SelectAdaptivePool2d(pool_type=global_pool)
        num_pooled_chs = head_chs * self.global_pool.feat_mult()
        self.conv_head = create_conv2d(num_pooled_chs, self.num_features, 1, padding=pad_type, bias=head_bias)
        self.act2 = act_layer(inplace=True)
        self.classifier = Linear(self.num_features, num_classes) if num_classes > 0 else nn.Identity()

        efficientnet_init_weights(self)

    def as_sequential(self):
        layers = [self.conv_stem, self.bn1, self.act1]
        layers.extend(self.blocks)
        layers.extend([self.global_pool, self.conv_head, self.act2])
        layers.extend([nn.Flatten(), nn.Dropout(self.drop_rate), self.classifier])
        return nn.Sequential(*layers)

    def get_classifier(self):
        return self.classifier

    def reset_classifier(self, num_classes, global_pool='avg'):
        self.num_classes = num_classes
        # cannot meaningfully change pooling of efficient head after creation
        self.global_pool = SelectAdaptivePool2d(pool_type=global_pool)
        self.classifier = Linear(self.num_features, num_classes) if num_classes > 0 else nn.Identity()

    def forward_features(self, x):
        x = self.conv_stem(x)
        x = self.bn1(x)
        x = self.act1(x)
        x = self.blocks(x)
        x = self.global_pool(x)
        x = self.conv_head(x)
        x = self.act2(x)
        return x

    def forward(self, x):
        x = self.forward_features(x)
        if not self.global_pool.is_identity():
            x = x.flatten(1)
        if self.drop_rate > 0.:
            x = F.dropout(x, p=self.drop_rate, training=self.training)
        return self.classifier(x)


class MobileNetV3Features(nn.Module):
    """ MobileNetV3 Feature Extractor

    A work-in-progress feature extraction module for MobileNet-V3 to use as a backbone for segmentation
    and object detection models.
    """

    def __init__(self, block_args, out_indices=(0, 1, 2, 3, 4), feature_location='bottleneck',
                 in_chans=3, stem_size=16, channel_multiplier=1.0, output_stride=32, pad_type='',
                 act_layer=nn.ReLU, drop_rate=0., drop_path_rate=0., se_kwargs=None,
                 norm_layer=nn.BatchNorm2d, norm_kwargs=None):
        super(MobileNetV3Features, self).__init__()
        norm_kwargs = norm_kwargs or {}
        self.drop_rate = drop_rate

        # Stem
        stem_size = round_channels(stem_size, channel_multiplier)
        self.conv_stem = create_conv2d(in_chans, stem_size, 3, stride=2, padding=pad_type)
        self.bn1 = norm_layer(stem_size, **norm_kwargs)
        self.act1 = act_layer(inplace=True)

        # Middle stages (IR/ER/DS Blocks)
        builder = EfficientNetBuilder(
            channel_multiplier, 8, None, output_stride, pad_type, act_layer, se_kwargs,
            norm_layer, norm_kwargs, drop_path_rate, feature_location=feature_location, verbose=_DEBUG)
        self.blocks = nn.Sequential(*builder(stem_size, block_args))
        self.feature_info = FeatureInfo(builder.features, out_indices)
        self._stage_out_idx = {v['stage']: i for i, v in enumerate(self.feature_info) if i in out_indices}

        efficientnet_init_weights(self)

        # Register feature extraction hooks with FeatureHooks helper
        self.feature_hooks = None
        if feature_location != 'bottleneck':
            hooks = self.feature_info.get_dicts(keys=('module', 'hook_type'))
            self.feature_hooks = FeatureHooks(hooks, self.named_modules())

    def forward(self, x) -> List[torch.Tensor]:
        x = self.conv_stem(x)
        x = self.bn1(x)
        x = self.act1(x)
        if self.feature_hooks is None:
            features = []
            if 0 in self._stage_out_idx:
                features.append(x)  # add stem out
            for i, b in enumerate(self.blocks):
                x = b(x)
                if i + 1 in self._stage_out_idx:
                    features.append(x)
            return features
        else:
            self.blocks(x)
            out = self.feature_hooks.get_output(x.device)
            return list(out.values())


def _create_mnv3(model_kwargs, variant, pretrained=False):
    features_only = False
    model_cls = MobileNetV3
    if model_kwargs.pop('features_only', False):
        features_only = True
        model_kwargs.pop('num_classes', 0)
        model_kwargs.pop('num_features', 0)
        model_kwargs.pop('head_conv', None)
        model_kwargs.pop('head_bias', None)
        model_cls = MobileNetV3Features
    model = build_model_with_cfg(
        model_cls, variant, pretrained, default_cfg=default_cfgs[variant],
        pretrained_strict=not features_only, **model_kwargs)
    if features_only:
        model.default_cfg = default_cfg_for_features(model.default_cfg)
    return model


def _gen_mobilenet_v3_rw(variant, channel_multiplier=1.0, pretrained=False, **kwargs):
    """Creates a MobileNet-V3 model.

    Ref impl: ?
    Paper: https://arxiv.org/abs/1905.02244

    Args:
      channel_multiplier: multiplier to number of channels per layer.
    """
    arch_def = [
        # stage 0, 112x112 in
        ['ds_r1_k3_s1_e1_c16_nre_noskip'],  # relu
        # stage 1, 112x112 in
        ['ir_r1_k3_s2_e4_c24_nre', 'ir_r1_k3_s1_e3_c24_nre'],  # relu
        # stage 2, 56x56 in
        ['ir_r3_k5_s2_e3_c40_se0.25_nre'],  # relu
        # stage 3, 28x28 in
        ['ir_r1_k3_s2_e6_c80', 'ir_r1_k3_s1_e2.5_c80', 'ir_r2_k3_s1_e2.3_c80'],  # hard-swish
        # stage 4, 14x14in
        ['ir_r2_k3_s1_e6_c112_se0.25'],  # hard-swish
        # stage 5, 14x14in
        ['ir_r3_k5_s2_e6_c160_se0.25'],  # hard-swish
        # stage 6, 7x7 in
        ['cn_r1_k1_s1_c960'],  # hard-swish
    ]
    model_kwargs = dict(
        block_args=decode_arch_def(arch_def),
        head_bias=False,
        channel_multiplier=channel_multiplier,
        norm_kwargs=resolve_bn_args(kwargs),
        act_layer=resolve_act_layer(kwargs, 'hard_swish'),
        se_kwargs=dict(gate_fn=get_act_fn('hard_sigmoid'), reduce_mid=True, divisor=1),
        **kwargs,
    )
    model = _create_mnv3(model_kwargs, variant, pretrained)
    return model


def _gen_mobilenet_v3(variant, channel_multiplier=1.0, pretrained=False, **kwargs):
    """Creates a MobileNet-V3 model.

    Ref impl: ?
    Paper: https://arxiv.org/abs/1905.02244

    Args:
      channel_multiplier: multiplier to number of channels per layer.
    """
    if 'small' in variant:
        num_features = 1024
        if 'minimal' in variant:
            act_layer = resolve_act_layer(kwargs, 'relu')
            arch_def = [
                # stage 0, 112x112 in
                ['ds_r1_k3_s2_e1_c16'],
                # stage 1, 56x56 in
                ['ir_r1_k3_s2_e4.5_c24', 'ir_r1_k3_s1_e3.67_c24'],
                # stage 2, 28x28 in
                ['ir_r1_k3_s2_e4_c40', 'ir_r2_k3_s1_e6_c40'],
                # stage 3, 14x14 in
                ['ir_r2_k3_s1_e3_c48'],
                # stage 4, 14x14in
                ['ir_r3_k3_s2_e6_c96'],
                # stage 6, 7x7 in
                ['cn_r1_k1_s1_c576'],
            ]
        else:
            act_layer = resolve_act_layer(kwargs, 'hard_swish')
            arch_def = [
                # stage 0, 112x112 in
                ['ds_r1_k3_s2_e1_c16_se0.25_nre'],  # relu
                # stage 1, 56x56 in
                ['ir_r1_k3_s2_e4.5_c24_nre', 'ir_r1_k3_s1_e3.67_c24_nre'],  # relu
                # stage 2, 28x28 in
                ['ir_r1_k5_s2_e4_c40_se0.25', 'ir_r2_k5_s1_e6_c40_se0.25'],  # hard-swish
                # stage 3, 14x14 in
                ['ir_r2_k5_s1_e3_c48_se0.25'],  # hard-swish
                # stage 4, 14x14in
                ['ir_r3_k5_s2_e6_c96_se0.25'],  # hard-swish
                # stage 6, 7x7 in
                ['cn_r1_k1_s1_c576'],  # hard-swish
            ]
    else:
        num_features = 1280
        if 'minimal' in variant:
            act_layer = resolve_act_layer(kwargs, 'relu')
            arch_def = [
                # stage 0, 112x112 in
                ['ds_r1_k3_s1_e1_c16'],
                # stage 1, 112x112 in
                ['ir_r1_k3_s2_e4_c24', 'ir_r1_k3_s1_e3_c24'],
                # stage 2, 56x56 in
                ['ir_r3_k3_s2_e3_c40'],
                # stage 3, 28x28 in
                ['ir_r1_k3_s2_e6_c80', 'ir_r1_k3_s1_e2.5_c80', 'ir_r2_k3_s1_e2.3_c80'],
                # stage 4, 14x14in
                ['ir_r2_k3_s1_e6_c112'],
                # stage 5, 14x14in
                ['ir_r3_k3_s2_e6_c160'],
                # stage 6, 7x7 in
                ['cn_r1_k1_s1_c960'],
            ]
        else:
            act_layer = resolve_act_layer(kwargs, 'hard_swish')
            arch_def = [
                # stage 0, 112x112 in
                ['ds_r1_k3_s1_e1_c16_nre'],  # relu
                # stage 1, 112x112 in
                ['ir_r1_k3_s2_e4_c24_nre', 'ir_r1_k3_s1_e3_c24_nre'],  # relu
                # stage 2, 56x56 in
                ['ir_r3_k5_s2_e3_c40_se0.25_nre'],  # relu
                # stage 3, 28x28 in
                ['ir_r1_k3_s2_e6_c80', 'ir_r1_k3_s1_e2.5_c80', 'ir_r2_k3_s1_e2.3_c80'],  # hard-swish
                # stage 4, 14x14in
                ['ir_r2_k3_s1_e6_c112_se0.25'],  # hard-swish
                # stage 5, 14x14in
                ['ir_r3_k5_s2_e6_c160_se0.25'],  # hard-swish
                # stage 6, 7x7 in
                ['cn_r1_k1_s1_c960'],  # hard-swish
            ]

    model_kwargs = dict(
        block_args=decode_arch_def(arch_def),
        num_features=num_features,
        stem_size=16,
        channel_multiplier=channel_multiplier,
        norm_kwargs=resolve_bn_args(kwargs),
        act_layer=act_layer,
        se_kwargs=dict(act_layer=nn.ReLU, gate_fn=hard_sigmoid, reduce_mid=True, divisor=8),
        **kwargs,
    )
    model = _create_mnv3(model_kwargs, variant, pretrained)
    return model


@register_model
def mobilenetv3_large_075(pretrained=False, **kwargs):
    """ MobileNet V3 """
    model = _gen_mobilenet_v3('mobilenetv3_large_075', 0.75, pretrained=pretrained, **kwargs)
    return model


@register_model
def mobilenetv3_large_100(pretrained=False, **kwargs):
    """ MobileNet V3 """
    model = _gen_mobilenet_v3('mobilenetv3_large_100', 1.0, pretrained=pretrained, **kwargs)
    return model


@register_model
def mobilenetv3_small_075(pretrained=False, **kwargs):
    """ MobileNet V3 """
    model = _gen_mobilenet_v3('mobilenetv3_small_075', 0.75, pretrained=pretrained, **kwargs)
    return model


@register_model
def mobilenetv3_small_100(pretrained=False, **kwargs):
    """ MobileNet V3 """
    model = _gen_mobilenet_v3('mobilenetv3_small_100', 1.0, pretrained=pretrained, **kwargs)
    return model


@register_model
def mobilenetv3_rw(pretrained=False, **kwargs):
    """ MobileNet V3 """
    if pretrained:
        # pretrained model trained with non-default BN epsilon
        kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
    model = _gen_mobilenet_v3_rw('mobilenetv3_rw', 1.0, pretrained=pretrained, **kwargs)
    return model


@register_model
def tf_mobilenetv3_large_075(pretrained=False, **kwargs):
    """ MobileNet V3 """
    kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
    kwargs['pad_type'] = 'same'
    model = _gen_mobilenet_v3('tf_mobilenetv3_large_075', 0.75, pretrained=pretrained, **kwargs)
    return model


@register_model
def tf_mobilenetv3_large_100(pretrained=False, **kwargs):
    """ MobileNet V3 """
    kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
    kwargs['pad_type'] = 'same'
    model = _gen_mobilenet_v3('tf_mobilenetv3_large_100', 1.0, pretrained=pretrained, **kwargs)
    return model


@register_model
def tf_mobilenetv3_large_minimal_100(pretrained=False, **kwargs):
    """ MobileNet V3 """
    kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
    kwargs['pad_type'] = 'same'
    model = _gen_mobilenet_v3('tf_mobilenetv3_large_minimal_100', 1.0, pretrained=pretrained, **kwargs)
    return model


@register_model
def tf_mobilenetv3_small_075(pretrained=False, **kwargs):
    """ MobileNet V3 """
    kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
    kwargs['pad_type'] = 'same'
    model = _gen_mobilenet_v3('tf_mobilenetv3_small_075', 0.75, pretrained=pretrained, **kwargs)
    return model


@register_model
def tf_mobilenetv3_small_100(pretrained=False, **kwargs):
    """ MobileNet V3 """
    kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
    kwargs['pad_type'] = 'same'
    model = _gen_mobilenet_v3('tf_mobilenetv3_small_100', 1.0, pretrained=pretrained, **kwargs)
    return model


@register_model
def tf_mobilenetv3_small_minimal_100(pretrained=False, **kwargs):
    """ MobileNet V3 """
    kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
    kwargs['pad_type'] = 'same'
    model = _gen_mobilenet_v3('tf_mobilenetv3_small_minimal_100', 1.0, pretrained=pretrained, **kwargs)
    return model