File size: 9,408 Bytes
a950ee6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
# Copyright (c) Meta Platforms, Inc. and affiliates.
# All rights reserved.

# This source code is licensed under the license found in the
# LICENSE file in the root directory of this source tree.

import torch
from torch import nn
from torch.nn import functional as F

from typing import List, Tuple, Type, Optional

from .common import LayerNorm2d

class MaskDecoder(nn.Module):
    def __init__(
        self,
        *,
        image_encoder_type: str,
        transformer_dim: int,
        transformer: nn.Module,
        num_multimask_outputs: int = 3,
        activation: Type[nn.Module] = nn.GELU,
        iou_head_depth: int = 3,
        iou_head_hidden_dim: int = 256,
        image_size,
        patch_size,
    ) -> None:
        """
        Predicts masks given an image and prompt embeddings, using a
        transformer architecture.

        Arguments:
          transformer_dim (int): the channel dimension of the transformer
          transformer (nn.Module): the transformer used to predict masks
          num_multimask_outputs (int): the number of masks to predict
            when disambiguating masks
          activation (nn.Module): the type of activation to use when
            upscaling masks
          iou_head_depth (int): the depth of the MLP used to predict
            mask quality
          iou_head_hidden_dim (int): the hidden dimension of the MLP
            used to predict mask quality
        """
        super().__init__()
        self.transformer_dim = transformer_dim
        self.transformer = transformer

        self.num_multimask_outputs = num_multimask_outputs

        self.iou_token = nn.Embedding(1, transformer_dim)
        self.num_mask_tokens = num_multimask_outputs + 1
        self.mask_tokens = nn.Embedding(self.num_mask_tokens, transformer_dim)

        if image_encoder_type == 'swin_vit':
            self.feat_shape = image_size/patch_size
            self.output_upscaling = nn.Sequential(
                nn.ConvTranspose3d(transformer_dim, transformer_dim // 4, kernel_size=2, stride=2),
                nn.LayerNorm((transformer_dim // 4, int(self.feat_shape[0]), int(self.feat_shape[1]), int(self.feat_shape[2]))),    # swin
                activation(),
                nn.ConvTranspose3d(transformer_dim // 4, transformer_dim // 8, kernel_size=2, stride=2),        # swin
                # nn.Conv3d(transformer_dim // 4, transformer_dim // 8, kernel_size=3, stride=1, padding=1),    # vit
                activation(),
            )
        else:
            self.feat_shape = image_size/patch_size * 2
            self.output_upscaling = nn.Sequential(
                nn.ConvTranspose3d(transformer_dim, transformer_dim // 4, kernel_size=2, stride=2),
                nn.LayerNorm((transformer_dim // 4, int(self.feat_shape[0]), int(self.feat_shape[1]), int(self.feat_shape[2]))), # vit
                activation(),
                nn.ConvTranspose3d(transformer_dim // 4, transformer_dim // 8, kernel_size=2, stride=2),
                # nn.Conv3d(transformer_dim // 4, transformer_dim // 8, kernel_size=3, stride=1, padding=1),
                activation(),
            )
        self.output_hypernetworks_mlps = nn.ModuleList(
            [
                MLP(transformer_dim, transformer_dim, transformer_dim // 8, 3)
                for i in range(self.num_mask_tokens)
            ]
        )

        self.iou_prediction_head = MLP(
            transformer_dim, iou_head_hidden_dim, self.num_mask_tokens, iou_head_depth
        )

        self.txt_align_upscaled_embedding = nn.Linear(768, 96)

    def forward(
        self,
        image_embeddings: torch.Tensor,
        text_embedding: Optional[torch.Tensor],
        image_pe: torch.Tensor,
        sparse_prompt_embeddings: torch.Tensor,
        dense_prompt_embeddings: torch.Tensor,
        multimask_output: bool,
    ) -> Tuple[torch.Tensor, torch.Tensor]:
        """
        Predict masks given image and prompt embeddings.

        Arguments:
          image_embeddings (torch.Tensor): the embeddings from the image encoder
          image_pe (torch.Tensor): positional encoding with the shape of image_embeddings
          sparse_prompt_embeddings (torch.Tensor): the embeddings of the points and boxes
          dense_prompt_embeddings (torch.Tensor): the embeddings of the mask inputs
          multimask_output (bool): Whether to return multiple masks or a single
            mask.

        Returns:
          torch.Tensor: batched predicted masks
          torch.Tensor: batched predictions of mask quality
        """
        # print('--------------decoder here--------------')
        masks, iou_pred = self.predict_masks(
            image_embeddings=image_embeddings,
            text_embedding=text_embedding,
            image_pe=image_pe,
            sparse_prompt_embeddings=sparse_prompt_embeddings,
            dense_prompt_embeddings=dense_prompt_embeddings,
        )

        # Select the correct mask or masks for output
        if multimask_output:
            mask_slice = slice(1, None)
        else:
            mask_slice = slice(0, 1)
        masks = masks[:, mask_slice, :, :, :]
        iou_pred = iou_pred[:, mask_slice]

        # Prepare output
        return masks, iou_pred

    def predict_masks(
        self,
        image_embeddings: torch.Tensor,
        text_embedding: torch.Tensor,
        image_pe: torch.Tensor,
        sparse_prompt_embeddings: torch.Tensor,
        dense_prompt_embeddings: torch.Tensor,
    ) -> Tuple[torch.Tensor, torch.Tensor]:
        """Predicts masks. See 'forward' for more details."""
        # Concatenate output tokens
        output_tokens = torch.cat([self.iou_token.weight, self.mask_tokens.weight], dim=0)
        output_tokens = output_tokens.unsqueeze(0).expand(sparse_prompt_embeddings.size(0), -1, -1)
        tokens = torch.cat((output_tokens, sparse_prompt_embeddings), dim=1)    # [2, 7=(5+2), 256]
        # Expand per-image data in batch direction to be per-mask
        if image_embeddings.shape[0] != tokens.shape[0]:
            src = torch.repeat_interleave(image_embeddings, tokens.shape[0], dim=0)
        else:
            src = image_embeddings
        src = src + dense_prompt_embeddings
        pos_src = torch.repeat_interleave(image_pe, tokens.shape[0], dim=0)
        b, c, h, w, d = src.shape

        # Run the transformer
        hs, src = self.transformer(src, pos_src, tokens)
        iou_token_out = hs[:, 0, :]
        mask_tokens_out = hs[:, 1 : (1 + self.num_mask_tokens), :]

        # Upscale mask embeddings and predict masks using the mask tokens
        src = src.transpose(1, 2).view(b, c, h, w, d)
        # print('src ', src.shape) # vit:[B, 768, 12, 12, 6], swin: [B, 6, 6, 3]
        upscaled_embedding = self.output_upscaling(src)
        hyper_in_list: List[torch.Tensor] = []
        for i in range(self.num_mask_tokens):
            hyper_in_list.append(self.output_hypernetworks_mlps[i](mask_tokens_out[:, i, :]))
        hyper_in = torch.stack(hyper_in_list, dim=1)
        b, c, h, w, d = upscaled_embedding.shape
        # print('hyper_in ', hyper_in.shape)    # [2, 4, 96]
        # print('upscaled_embedding ', upscaled_embedding.shape)    # [2, 96, 24, 24, 12]*
        masks = (hyper_in @ upscaled_embedding.view(b, c, h * w * d)).view(b, -1, h, w, d)
        # print('masks here ', masks.shape) # [2, 4, 24, 24, 12]
        
        if text_embedding is not None:
            # text_embedding: B x 768, upscaled_embedding: B x c x h x w x d => B x 1 x h x w x d
            text_embedding_down = self.txt_align_upscaled_embedding(text_embedding).unsqueeze(dim=1)
            upscaled_embedding = upscaled_embedding.view(b, c, h * w * d)
            # print('text_embedding_down ', text_embedding_down.shape)  # [2, 1, 96]
            # text_embedding_norm = F.normalize(text_embedding_down, dim=-1)
            # upscaled_embedding_norm = F.normalize(upscaled_embedding, dim=1)
            # sim = (text_embedding_norm @ upscaled_embedding_norm).view(b, -1, h, w, d)
            # print(text_embedding_down.shape, upscaled_embedding.shape)
            sim = (text_embedding_down @ upscaled_embedding).view(b, -1, h, w, d)
            # print('sim ', sim.shape)  # [B, 1, 24, 24, 12]
            sim = sim.repeat(1, masks.shape[1], 1, 1, 1)
            # print('sim after', sim.shape) # [B, 4, 24, 24, 12]
            masks = masks + sim
        # Generate mask quality predictions
        iou_pred = self.iou_prediction_head(iou_token_out)

        return masks, iou_pred


# Lightly adapted from
# https://github.com/facebookresearch/MaskFormer/blob/main/mask_former/modeling/transformer/transformer_predictor.py # noqa
class MLP(nn.Module):
    def __init__(
        self,
        input_dim: int,
        hidden_dim: int,
        output_dim: int,
        num_layers: int,
        sigmoid_output: bool = False,
    ) -> None:
        super().__init__()
        self.num_layers = num_layers
        h = [hidden_dim] * (num_layers - 1)
        self.layers = nn.ModuleList(
            nn.Linear(n, k) for n, k in zip([input_dim] + h, h + [output_dim])
        )
        self.sigmoid_output = sigmoid_output

    def forward(self, x):
        for i, layer in enumerate(self.layers):
            x = F.relu(layer(x)) if i < self.num_layers - 1 else layer(x)
        if self.sigmoid_output:
            x = F.sigmoid(x)
        return x