[models] update frame-level feature extraction interface (#367)

wespeaker/models/ecapa_tdnn.py

@@ -205,7 +205,7 @@ def __init__(self,
         else:
             self.bn2 = nn.Identity()
 
-    def __get_frame_level_feat(self, x):
+    def _get_frame_level_feat(self, x):
         # for inner class usage
         x = x.permute(0, 2, 1)  # (B,T,F) -> (B,F,T)
 
@@ -221,11 +221,11 @@ def __get_frame_level_feat(self, x):
 
     def get_frame_level_feat(self, x):
         # for outer interface
-        out = self.__get_frame_level_feat(x).permute(0, 2, 1)
+        out = self._get_frame_level_feat(x).permute(0, 2, 1)
         return out  # (B, T, D)
 
     def forward(self, x):
-        out = F.relu(self.__get_frame_level_feat(x))
+        out = F.relu(self._get_frame_level_feat(x))
         out = self.bn(self.pool(out))
         out = self.linear(out)
         if self.emb_bn:

wespeaker/models/eres2net.py

@@ -351,7 +351,7 @@ def _make_layer(self,
             self.in_planes = planes * self.expansion
         return nn.Sequential(*layers)
 
-    def __get_frame_level_feat(self, x):
+    def _get_frame_level_feat(self, x):
         # for inner class usage
         x = x.permute(0, 2, 1)  # (B,T,F) => (B,F,T)
         x = x.unsqueeze_(1)
@@ -371,14 +371,14 @@ def __get_frame_level_feat(self, x):
 
     def get_frame_level_feat(self, x):
         # for outer interface
-        out = self.__get_frame_level_feat(x)
+        out = self._get_frame_level_feat(x)
         out = out.transpose(1, 3)
         out = torch.flatten(out, 2, -1)
 
         return out  # (B, T, D)
 
     def forward(self, x):
-        fuse_out1234 = self.__get_frame_level_feat(x)
+        fuse_out1234 = self._get_frame_level_feat(x)
         stats = self.pool(fuse_out1234)
 
         embed_a = self.seg_1(stats)

wespeaker/models/gemini_dfresnet.py

@@ -102,7 +102,7 @@ def __init__(self,
             self.seg_bn_1 = nn.Identity()
             self.seg_2 = nn.Identity()
 
-    def __get_frame_level_feat(self, x):
+    def _get_frame_level_feat(self, x):
         # for inner class usage
         x = x.permute(0, 2, 1)  # (B,T,F) => (B,F,T)
         x = x.unsqueeze_(1)
@@ -120,15 +120,15 @@ def __get_frame_level_feat(self, x):
 
     def get_frame_level_feat(self, x):
         # for outer interface 
-        out = self.__get_frame_level_feat(x)
+        out = self._get_frame_level_feat(x)
         out = out.transpose(1, 3)
         out = torch.flatten(out, 2, -1)
 
         return out  # (B, T, D)
 
     def forward(self, x):
 
-        out = self.__get_frame_level_feat(x)
+        out = self._get_frame_level_feat(x)
         stats = self.pool(out)
 
         embed_a = self.seg_1(stats)

wespeaker/models/redimnet.py

@@ -843,7 +843,7 @@ def __init__(
             self.seg_bn_1 = nn.Identity()
             self.seg_2 = nn.Identity()
 
-    def __get_frame_level_feat(self, x):
+    def _get_frame_level_feat(self, x):
         # for inner class usage
         x = x.permute(0, 2, 1)  # (B,F,T) => (B,T,F)
         x = x.unsqueeze_(1)
@@ -853,12 +853,12 @@ def __get_frame_level_feat(self, x):
 
     def get_frame_level_feat(self, x):
         # for outer interface
-        out = self.__get_frame_level_feat(x).permute(0, 2, 1)
+        out = self._get_frame_level_feat(x).permute(0, 2, 1)
 
         return out  # (B, T, D)
 
     def forward(self, x):
-        out = self.__get_frame_level_feat(x)
+        out = self._get_frame_level_feat(x)
 
         stats = self.pool(out)
         embed_a = self.seg_1(stats)

wespeaker/models/repvgg.py

@@ -559,7 +559,7 @@ def get_downsample_multiple(self):
     def get_output_planes(self):
         return self.output_planes
 
-    def __get_frame_level_feat(self, x):
+    def _get_frame_level_feat(self, x):
         # for inner class usage
         x = x.permute(0, 2, 1)  # (B,T,F) -> (B,F,T)
         x = x.unsqueeze_(1)
@@ -573,14 +573,14 @@ def __get_frame_level_feat(self, x):
 
     def get_frame_level_feat(self, x):
         # for outer interface
-        out = self.__get_frame_level_feat(x)
+        out = self._get_frame_level_feat(x)
         out = out.transpose(1, 3)
         out = torch.flatten(out, 2, -1)
 
         return out  # (B, T, D)
 
     def forward(self, x):
-        x = self.__get_frame_level_feat(x)
+        x = self._get_frame_level_feat(x)
         stats = self.pool(x)
         embed = self.seg(stats)
 

wespeaker/models/res2net.py

@@ -154,7 +154,7 @@ def _make_layer(self, block, planes, num_blocks, stride):
             self.in_planes = planes * block.expansion
         return nn.Sequential(*layers)
 
-    def __get_frame_level_feat(self, x):
+    def _get_frame_level_feat(self, x):
         # for inner class usage
         x = x.permute(0, 2, 1)  # (B,T,F) => (B,F,T)
 
@@ -169,14 +169,14 @@ def __get_frame_level_feat(self, x):
 
     def get_frame_level_feat(self, x):
         # for outer interface
-        out = self.__get_frame_level_feat(x)
+        out = self._get_frame_level_feat(x)
         out = out.transpose(1, 3)
         out = torch.flatten(out, 2, -1)
 
         return out  # (B, T, D)
 
     def forward(self, x):
-        out = self.__get_frame_level_feat(x)
+        out = self._get_frame_level_feat(x)
         stats = self.pool(out)
 
         embed_a = self.seg_1(stats)

wespeaker/models/resnet.py

@@ -168,7 +168,7 @@ def _make_layer(self, block, planes, num_blocks, stride):
             self.in_planes = planes * block.expansion
         return nn.Sequential(*layers)
 
-    def __get_frame_level_feat(self, x):
+    def _get_frame_level_feat(self, x):
         # for inner class usage
         x = x.permute(0, 2, 1)  # (B,T,F) => (B,F,T)
 
@@ -183,14 +183,14 @@ def __get_frame_level_feat(self, x):
 
     def get_frame_level_feat(self, x):
         # for outer interface
-        out = self.__get_frame_level_feat(x)
+        out = self._get_frame_level_feat(x)
         out = out.transpose(1, 3)
         out = torch.flatten(out, 2, -1)
 
         return out  # (B, T, D)
 
     def forward(self, x):
-        out = self.__get_frame_level_feat(x)
+        out = self._get_frame_level_feat(x)
 
         stats = self.pool(out)
 

wespeaker/models/tdnn.py

@@ -86,7 +86,7 @@ def __init__(self,
         self.seg_bn_1 = nn.BatchNorm1d(embed_dim, affine=False)
         self.seg_2 = nn.Linear(embed_dim, embed_dim)
 
-    def __get_frame_level_feat(self, x):
+    def _get_frame_level_feat(self, x):
         # for inner class usage
         x = x.permute(0, 2, 1)  # (B,T,F) -> (B,F,T)
 
@@ -100,12 +100,12 @@ def __get_frame_level_feat(self, x):
 
     def get_frame_level_feat(self, x):
         # for outer interface
-        out = self.__get_frame_level_feat(x).permute(0, 2, 1)
+        out = self._get_frame_level_feat(x).permute(0, 2, 1)
 
         return out  # (B, T, D)
 
     def forward(self, x):
-        out = self.__get_frame_level_feat(x)
+        out = self._get_frame_level_feat(x)
         stats = self.pool(out)
         embed_a = self.seg_1(stats)
         out = F.relu(embed_a)