Fix `NaiveSyncBatchNorm1d` and `NaiveSyncBatchNorm2d` (#1435)

* add quick install command

* fix SyncBatchNorm

* fix SyncBatchNorm

mmdet3d/ops/norm.py

@@ -53,11 +53,27 @@ class NaiveSyncBatchNorm1d(nn.BatchNorm1d):
     # TODO: make mmcv fp16 utils handle customized norm layers
     @force_fp32(out_fp16=True)
     def forward(self, input):
+        """
+        Args:
+            input (tensor): Has shape (N, C) or (N, C, L), where N is
+                the batch size, C is the number of features or
+                channels, and L is the sequence length
+
+        Returns:
+            tensor: Has shape (N, C) or (N, C, L), has same shape
+            as input.
+        """
         assert input.dtype == torch.float32, \
             f'input should be in float32 type, got {input.dtype}'
-        if dist.get_world_size() == 1 or not self.training:
+        using_dist = dist.is_available() and dist.is_initialized()
+        if (not using_dist) or dist.get_world_size() == 1 \
+                or not self.training:
             return super().forward(input)
         assert input.shape[0] > 0, 'SyncBN does not support empty inputs'
+        is_two_dim = input.dim() == 2
+        if is_two_dim:
+            input = input.unsqueeze(2)
+
         C = input.shape[1]
         mean = torch.mean(input, dim=[0, 2])
         meansqr = torch.mean(input * input, dim=[0, 2])
@@ -76,7 +92,10 @@ class NaiveSyncBatchNorm1d(nn.BatchNorm1d):
         bias = self.bias - mean * scale
         scale = scale.reshape(1, -1, 1)
         bias = bias.reshape(1, -1, 1)
-        return input * scale + bias
+        output = input * scale + bias
+        if is_two_dim:
+            output = output.squeeze(2)
+        return output
 
 
 @NORM_LAYERS.register_module('naiveSyncBN2d')
@@ -107,9 +126,19 @@ class NaiveSyncBatchNorm2d(nn.BatchNorm2d):
     # TODO: make mmcv fp16 utils handle customized norm layers
     @force_fp32(out_fp16=True)
     def forward(self, input):
+        """
+        Args:
+            Input (tensor): Feature has shape (N, C, H, W).
+
+        Returns:
+            tensor: Has shape (N, C, H, W), same shape as input.
+        """
         assert input.dtype == torch.float32, \
             f'input should be in float32 type, got {input.dtype}'
-        if dist.get_world_size() == 1 or not self.training:
+        using_dist = dist.is_available() and dist.is_initialized()
+        if (not using_dist) or \
+                dist.get_world_size() == 1 or \
+                not self.training:
             return super().forward(input)
 
         assert input.shape[0] > 0, 'SyncBN does not support empty inputs'