take AdaptiveAvgPoolnd to replace origin one

config/i3d_rgb_32x2x1_r50_3d_kinetics400_100e.py

@@ -14,8 +14,6 @@ model = dict(
         num_classes=400,
         in_channels=2048,
         spatial_type='avg',
-        spatial_size=7,
-        temporal_size=4,
         dropout_ratio=0.5,
         init_std=0.01))
 # model training and testing settings

config/tsn_rgb_1x1x3_r50_2d_kinetics400_100e.py

@@ -11,7 +11,6 @@ model = dict(
         num_classes=400,
         in_channels=2048,
         spatial_type='avg',
-        spatial_size=7,
         consensus=dict(type='AvgConsensus', dim=1),
         dropout_ratio=0.4,
         init_std=0.01))

mmaction/models/heads/i3d_head.py

-import mmcv
 import torch.nn as nn
 from mmcv.cnn.weight_init import normal_init
-from torch.nn.modules.utils import _pair
 
 from ..registry import HEADS
 from .base import BaseHead
@@ -27,17 +25,11 @@ class I3DHead(BaseHead):
                  num_classes,
                  in_channels=2048,
                  spatial_type='avg',
-                 spatial_size=7,
-                 temporal_size=4,
                  dropout_ratio=0.5,
                  init_std=0.01):
         super(I3DHead, self).__init__(num_classes, in_channels)
-        self.spatial_size = _pair(spatial_size)
-        assert mmcv.is_tuple_of(self.spatial_size, int)
 
         self.spatial_type = spatial_type
-        self.temporal_size = temporal_size
-        self.pool_size = (self.temporal_size, ) + self.spatial_size
         self.dropout_ratio = dropout_ratio
         self.init_std = init_std
         if self.dropout_ratio != 0:
@@ -47,7 +39,8 @@ class I3DHead(BaseHead):
         self.fc_cls = nn.Linear(self.in_channels, self.num_classes)
 
         if self.spatial_type == 'avg':
-            self.avg_pool = nn.AvgPool3d(self.pool_size, stride=1, padding=0)
+            # use `nn.AdaptiveAvgPool3d` to adaptively match the in_channels.
+            self.avg_pool = nn.AdaptiveAvgPool3d((1, 1, 1))
         else:
             self.avg_pool = None
 

mmaction/models/heads/tsn_head.py

-import mmcv
 import torch.nn as nn
 from mmcv.cnn.weight_init import normal_init
-from torch.nn.modules.utils import _pair
 
 from ..registry import HEADS
 from .base import BaseHead
@@ -41,13 +39,10 @@ class TSNHead(BaseHead):
                  num_classes,
                  in_channels=2048,
                  spatial_type='avg',
-                 spatial_size=7,
                  consensus=dict(type='AvgConsensus', dim=1),
                  dropout_ratio=0.4,
                  init_std=0.01):
         super(TSNHead, self).__init__(num_classes, in_channels)
-        self.spatial_size = _pair(spatial_size)
-        assert mmcv.is_tuple_of(self.spatial_size, int)
 
         self.spatial_type = spatial_type
         self.dropout_ratio = dropout_ratio
@@ -60,10 +55,10 @@ class TSNHead(BaseHead):
             self.consensus = None
 
         if self.spatial_type == 'avg':
-            self.avg_pool2d = nn.AvgPool2d(
-                self.spatial_size, stride=1, padding=0)
+            # use `nn.AdaptiveAvgPool2d` to adaptively match the in_channels.
+            self.avg_pool = nn.AdaptiveAvgPool2d((1, 1))
         else:
-            self.avg_pool2d = None
+            self.avg_pool = None
 
         if self.dropout_ratio != 0:
             self.dropout = nn.Dropout(p=self.dropout_ratio)
@@ -76,7 +71,7 @@ class TSNHead(BaseHead):
 
     def forward(self, x, num_segs):
         # [N * num_segs, in_channels, 7, 7]
-        x = self.avg_pool2d(x)
+        x = self.avg_pool(x)
         # [N * num_segs, in_channels, 1, 1]
         x = x.reshape((-1, num_segs) + x.shape[1:])
         # [N, num_segs, in_channels, 1, 1]