Use pytorch photometric loss instead of old self-rolled

model/networks.py

@@ -387,7 +387,7 @@ class RectifiedPatternSimilarityLoss(TimedModule):
         if std is not None:
             mask = mask * std
 
-        diff = torchext.photometric_loss(pattern_proj.contiguous(), im.contiguous(), 9, self.loss_type, self.loss_eps)
+        diff = torchext.photometric_loss_pytorch(pattern_proj.contiguous(), im.contiguous(), 9, self.loss_type, self.loss_eps)
         val = (mask * diff).sum() / mask.sum()
         return val, pattern_proj