File size: 6,935 Bytes
9bf4bd7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
# Copyright (c) OpenMMLab. All rights reserved.
from typing import Tuple, Union

import torch
import torch.nn as nn

from mmocr.registry import MODELS


@MODELS.register_module()
class Maxpool2d(nn.Module):
    """A wrapper around nn.Maxpool2d().

    Args:
        kernel_size (int or tuple(int)): Kernel size for max pooling layer
        stride (int or tuple(int)): Stride for max pooling layer
        padding (int or tuple(int)): Padding for pooling layer
    """

    def __init__(self,
                 kernel_size: Union[int, Tuple[int]],
                 stride: Union[int, Tuple[int]],
                 padding: Union[int, Tuple[int]] = 0,
                 **kwargs) -> None:
        super().__init__()
        self.model = nn.MaxPool2d(kernel_size, stride, padding)

    def forward(self, x) -> torch.Tensor:
        """Forward function.
        Args:
            x (Tensor): Input feature map.

        Returns:
            Tensor: Output tensor after Maxpooling layer.
        """
        return self.model(x)


@MODELS.register_module()
class GCAModule(nn.Module):
    """GCAModule in MASTER.

    Args:
        in_channels (int): Channels of input tensor.
        ratio (float): Scale ratio of in_channels.
        n_head (int): Numbers of attention head.
        pooling_type (str): Spatial pooling type. Options are [``avg``,
            ``att``].
        scale_attn (bool): Whether to scale the attention map. Defaults to
            False.
        fusion_type (str): Fusion type of input and context. Options are
            [``channel_add``, ``channel_mul``, ``channel_concat``].
    """

    def __init__(self,
                 in_channels: int,
                 ratio: float,
                 n_head: int,
                 pooling_type: str = 'att',
                 scale_attn: bool = False,
                 fusion_type: str = 'channel_add',
                 **kwargs) -> None:
        super().__init__()

        assert pooling_type in ['avg', 'att']
        assert fusion_type in ['channel_add', 'channel_mul', 'channel_concat']

        # in_channels must be divided by headers evenly
        assert in_channels % n_head == 0 and in_channels >= 8

        self.n_head = n_head
        self.in_channels = in_channels
        self.ratio = ratio
        self.planes = int(in_channels * ratio)
        self.pooling_type = pooling_type
        self.fusion_type = fusion_type
        self.scale_attn = scale_attn
        self.single_header_inplanes = int(in_channels / n_head)

        if pooling_type == 'att':
            self.conv_mask = nn.Conv2d(
                self.single_header_inplanes, 1, kernel_size=1)
            self.softmax = nn.Softmax(dim=2)
        else:
            self.avg_pool = nn.AdaptiveAvgPool2d(1)

        if fusion_type == 'channel_add':
            self.channel_add_conv = nn.Sequential(
                nn.Conv2d(self.in_channels, self.planes, kernel_size=1),
                nn.LayerNorm([self.planes, 1, 1]), nn.ReLU(inplace=True),
                nn.Conv2d(self.planes, self.in_channels, kernel_size=1))
        elif fusion_type == 'channel_concat':
            self.channel_concat_conv = nn.Sequential(
                nn.Conv2d(self.in_channels, self.planes, kernel_size=1),
                nn.LayerNorm([self.planes, 1, 1]), nn.ReLU(inplace=True),
                nn.Conv2d(self.planes, self.in_channels, kernel_size=1))
            # for concat
            self.cat_conv = nn.Conv2d(
                2 * self.in_channels, self.in_channels, kernel_size=1)
        elif fusion_type == 'channel_mul':
            self.channel_mul_conv = nn.Sequential(
                nn.Conv2d(self.in_channels, self.planes, kernel_size=1),
                nn.LayerNorm([self.planes, 1, 1]), nn.ReLU(inplace=True),
                nn.Conv2d(self.planes, self.in_channels, kernel_size=1))

    def spatial_pool(self, x: torch.Tensor) -> torch.Tensor:
        """Spatial pooling function.

        Args:
            x (Tensor): Input feature map.

        Returns:
            Tensor: Output tensor after spatial pooling.
        """
        batch, channel, height, width = x.size()
        if self.pooling_type == 'att':
            # [N*headers, C', H , W] C = headers * C'
            x = x.view(batch * self.n_head, self.single_header_inplanes,
                       height, width)
            input_x = x

            # [N*headers, C', H * W] C = headers * C'
            input_x = input_x.view(batch * self.n_head,
                                   self.single_header_inplanes, height * width)

            # [N*headers, 1, C', H * W]
            input_x = input_x.unsqueeze(1)
            # [N*headers, 1, H, W]
            context_mask = self.conv_mask(x)
            # [N*headers, 1, H * W]
            context_mask = context_mask.view(batch * self.n_head, 1,
                                             height * width)

            # scale variance
            if self.scale_attn and self.n_head > 1:
                context_mask = context_mask / \
                               torch.sqrt(self.single_header_inplanes)

            # [N*headers, 1, H * W]
            context_mask = self.softmax(context_mask)

            # [N*headers, 1, H * W, 1]
            context_mask = context_mask.unsqueeze(-1)
            # [N*headers, 1, C', 1] =
            # [N*headers, 1, C', H * W] * [N*headers, 1, H * W, 1]
            context = torch.matmul(input_x, context_mask)

            # [N, headers * C', 1, 1]
            context = context.view(batch,
                                   self.n_head * self.single_header_inplanes,
                                   1, 1)
        else:
            # [N, C, 1, 1]
            context = self.avg_pool(x)

        return context

    def forward(self, x: torch.Tensor) -> torch.Tensor:
        """Forward function.

        Args:
            x (Tensor): Input feature map.

        Returns:
            Tensor: Output tensor after GCAModule.
        """
        # [N, C, 1, 1]
        context = self.spatial_pool(x)
        out = x

        if self.fusion_type == 'channel_mul':
            # [N, C, 1, 1]
            channel_mul_term = torch.sigmoid(self.channel_mul_conv(context))
            out = out * channel_mul_term
        elif self.fusion_type == 'channel_add':
            # [N, C, 1, 1]
            channel_add_term = self.channel_add_conv(context)
            out = out + channel_add_term
        else:
            # [N, C, 1, 1]
            channel_concat_term = self.channel_concat_conv(context)

            # use concat
            _, C1, _, _ = channel_concat_term.shape
            N, C2, H, W = out.shape

            out = torch.cat([out,
                             channel_concat_term.expand(-1, -1, H, W)],
                            dim=1)
            out = self.cat_conv(out)
            out = nn.functional.layer_norm(out, [self.in_channels, H, W])
            out = nn.functional.relu(out)

        return out