update pefd

Author: user3984

ppcls/loss/pefdloss.py

@@ -24,7 +24,7 @@ class Regressor(nn.Layer):
 
     def __init__(self, dim_in=1024, dim_out=1024):
         super(Regressor, self).__init__()
-        self.conv = nn.Conv2D(dim_in, dim_out, 1)
+        self.conv = nn.Linear(dim_in, dim_out)
 
     def forward(self, x):
         x = self.conv(x)
@@ -38,29 +38,38 @@ class PEFDLoss(nn.Layer):
     Code reference: https://github.com/chenyd7/PEFD
     """
 
-    def __init__(self, student_channel, teacher_channel, num_projectors=3):
+    def __init__(self,
+                 student_channel,
+                 teacher_channel,
+                 num_projectors=3,
+                 mode="flatten"):
         super().__init__()
 
         if num_projectors <= 0:
             raise ValueError("Number of projectors must be greater than 0.")
 
+        if mode not in ["flatten", "gap"]:
+            raise ValueError("Mode must be \"flatten\" or \"gap\".")
+
+        self.mode = mode
         self.projectors = nn.LayerList()
 
         for _ in range(num_projectors):
             self.projectors.append(Regressor(student_channel, teacher_channel))
 
     def forward(self, student_feature, teacher_feature):
-        if student_feature.shape[2:] != teacher_feature.shape[2:]:
-            raise ValueError(
-                "Student feature must have the same H and W as teacher feature."
-            )
+        if self.mode == "gap":
+            student_feature = F.adaptive_avg_pool2d(student_feature, (1, 1))
+            teacher_feature = F.adaptive_avg_pool2d(teacher_feature, (1, 1))
+
+        student_feature = student_feature.flatten(1)
+        f_t = teacher_feature.flatten(1)
 
         q = len(self.projectors)
         f_s = 0.0
         for i in range(q):
             f_s += self.projectors[i](student_feature)
-        f_s = (f_s / q).flatten(1)
-        f_t = teacher_feature.flatten(1)
+        f_s = f_s / q
 
         # inner product (normalize first and inner product)
         normft = f_t.pow(2).sum(1, keepdim=True).pow(1. / 2)