File size: 9,369 Bytes
d670799
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
# Copyright (c) OpenMMLab. All rights reserved.
from typing import Dict, List, Tuple, Union

import torch
import torch.nn as nn
import torch.nn.functional as F
from mmengine.model.weight_init import normal_init

from mmaction.evaluation import top_k_accuracy
from mmaction.registry import MODELS
from mmaction.utils import SampleList
from .base import BaseHead


@MODELS.register_module()
class RGBPoseHead(BaseHead):
    """The classification head for RGBPoseConv3D.



    Args:

        num_classes (int): Number of classes to be classified.

        in_channels (tuple[int]): Number of channels in input feature.

        loss_cls (dict): Config for building loss.

            Defaults to ``dict(type='CrossEntropyLoss')``.

        loss_components (list[str]): The components of the loss.

            Defaults to ``['rgb', 'pose']``.

        loss_weights (float or tuple[float]): The weights of the losses.

            Defaults to 1.

        dropout (float): Probability of dropout layer. Default: 0.5.

        init_std (float): Std value for Initiation. Default: 0.01.

    """

    def __init__(self,

                 num_classes: int,

                 in_channels: Tuple[int],

                 loss_cls: Dict = dict(type='CrossEntropyLoss'),

                 loss_components: List[str] = ['rgb', 'pose'],

                 loss_weights: Union[float, Tuple[float]] = 1.,

                 dropout: float = 0.5,

                 init_std: float = 0.01,

                 **kwargs) -> None:
        super().__init__(num_classes, in_channels, loss_cls, **kwargs)
        if isinstance(dropout, float):
            dropout = {'rgb': dropout, 'pose': dropout}
        assert isinstance(dropout, dict)

        if loss_components is not None:
            self.loss_components = loss_components
            if isinstance(loss_weights, float):
                loss_weights = [loss_weights] * len(loss_components)
            assert len(loss_weights) == len(loss_components)
            self.loss_weights = loss_weights

        self.dropout = dropout
        self.init_std = init_std

        self.dropout_rgb = nn.Dropout(p=self.dropout['rgb'])
        self.dropout_pose = nn.Dropout(p=self.dropout['pose'])

        self.fc_rgb = nn.Linear(self.in_channels[0], num_classes)
        self.fc_pose = nn.Linear(self.in_channels[1], num_classes)
        self.avg_pool = nn.AdaptiveAvgPool3d((1, 1, 1))

    def init_weights(self) -> None:
        """Initiate the parameters from scratch."""
        normal_init(self.fc_rgb, std=self.init_std)
        normal_init(self.fc_pose, std=self.init_std)

    def forward(self, x: Tuple[torch.Tensor]) -> Dict:
        """Defines the computation performed at every call."""
        x_rgb, x_pose = self.avg_pool(x[0]), self.avg_pool(x[1])
        x_rgb = x_rgb.view(x_rgb.size(0), -1)
        x_pose = x_pose.view(x_pose.size(0), -1)

        x_rgb = self.dropout_rgb(x_rgb)
        x_pose = self.dropout_pose(x_pose)

        cls_scores = dict()
        cls_scores['rgb'] = self.fc_rgb(x_rgb)
        cls_scores['pose'] = self.fc_pose(x_pose)

        return cls_scores

    def loss(self, feats: Tuple[torch.Tensor], data_samples: SampleList,

             **kwargs) -> Dict:
        """Perform forward propagation of head and loss calculation on the

        features of the upstream network.



        Args:

            feats (tuple[torch.Tensor]): Features from upstream network.

            data_samples (list[:obj:`ActionDataSample`]): The batch

                data samples.



        Returns:

            dict: A dictionary of loss components.

        """
        cls_scores = self(feats, **kwargs)
        return self.loss_by_feat(cls_scores, data_samples)

    def loss_by_feat(self, cls_scores: Dict[str, torch.Tensor],

                     data_samples: SampleList) -> Dict:
        """Calculate the loss based on the features extracted by the head.



        Args:

            cls_scores (dict[str, torch.Tensor]): The dict of

                classification scores,

            data_samples (list[:obj:`ActionDataSample`]): The batch

                data samples.



        Returns:

            dict: A dictionary of loss components.

        """
        labels = torch.stack([x.gt_label for x in data_samples])
        labels = labels.squeeze()

        if labels.shape == torch.Size([]):
            labels = labels.unsqueeze(0)
        elif labels.dim() == 1 and labels.size()[0] == self.num_classes \
                and cls_scores.size()[0] == 1:
            # Fix a bug when training with soft labels and batch size is 1.
            # When using soft labels, `labels` and `cls_score` share the same
            # shape.
            labels = labels.unsqueeze(0)

        losses = dict()
        for loss_name, weight in zip(self.loss_components, self.loss_weights):
            cls_score = cls_scores[loss_name]
            loss_cls = self.loss_by_scores(cls_score, labels)
            loss_cls = {loss_name + '_' + k: v for k, v in loss_cls.items()}
            loss_cls[f'{loss_name}_loss_cls'] *= weight
            losses.update(loss_cls)
        return losses

    def loss_by_scores(self, cls_scores: torch.Tensor,

                       labels: torch.Tensor) -> Dict:
        """Calculate the loss based on the features extracted by the head.



        Args:

            cls_scores (torch.Tensor): Classification prediction

                results of all class, has shape (batch_size, num_classes).

            labels (torch.Tensor): The labels used to calculate the loss.



        Returns:

            dict: A dictionary of loss components.

        """
        losses = dict()
        if cls_scores.size() != labels.size():
            top_k_acc = top_k_accuracy(cls_scores.detach().cpu().numpy(),
                                       labels.detach().cpu().numpy(),
                                       self.topk)
            for k, a in zip(self.topk, top_k_acc):
                losses[f'top{k}_acc'] = torch.tensor(
                    a, device=cls_scores.device)
        if self.label_smooth_eps != 0:
            if cls_scores.size() != labels.size():
                labels = F.one_hot(labels, num_classes=self.num_classes)
            labels = ((1 - self.label_smooth_eps) * labels +
                      self.label_smooth_eps / self.num_classes)

        loss_cls = self.loss_cls(cls_scores, labels)
        # loss_cls may be dictionary or single tensor
        if isinstance(loss_cls, dict):
            losses.update(loss_cls)
        else:
            losses['loss_cls'] = loss_cls
        return losses

    def predict(self, feats: Tuple[torch.Tensor], data_samples: SampleList,

                **kwargs) -> SampleList:
        """Perform forward propagation of head and predict recognition results

        on the features of the upstream network.



        Args:

            feats (tuple[torch.Tensor]): Features from upstream network.

            data_samples (list[:obj:`ActionDataSample`]): The batch

                data samples.



        Returns:

             list[:obj:`ActionDataSample`]: Recognition results wrapped

                by :obj:`ActionDataSample`.

        """
        cls_scores = self(feats, **kwargs)
        return self.predict_by_feat(cls_scores, data_samples)

    def predict_by_feat(self, cls_scores: Dict[str, torch.Tensor],

                        data_samples: SampleList) -> SampleList:
        """Transform a batch of output features extracted from the head into

        prediction results.



        Args:

            cls_scores (dict[str, torch.Tensor]): The dict of

                classification scores,

            data_samples (list[:obj:`ActionDataSample`]): The

                annotation data of every samples. It usually includes

                information such as `gt_label`.



        Returns:

            list[:obj:`ActionDataSample`]: Recognition results wrapped

                by :obj:`ActionDataSample`.

        """
        pred_scores = [dict() for _ in range(len(data_samples))]

        for name in self.loss_components:
            cls_score = cls_scores[name]
            cls_score = self.predict_by_scores(cls_score, data_samples)
            for pred_score, score in zip(pred_scores, cls_score):
                pred_score[f'{name}'] = score

        for data_sample, pred_score, in zip(data_samples, pred_scores):
            data_sample.set_pred_score(pred_score)
        return data_samples

    def predict_by_scores(self, cls_scores: torch.Tensor,

                          data_samples: SampleList) -> torch.Tensor:
        """Transform a batch of output features extracted from the head into

        prediction results.



        Args:

            cls_scores (torch.Tensor): Classification scores, has a shape

                (B*num_segs, num_classes)

            data_samples (list[:obj:`ActionDataSample`]): The annotation

                data of every samples.



        Returns:

            torch.Tensor: The averaged classification scores.

        """

        num_segs = cls_scores.shape[0] // len(data_samples)
        cls_scores = self.average_clip(cls_scores, num_segs=num_segs)
        return cls_scores