[BC-Breaking] Replace waveform with specgram in SlidingWindowCmn (#1859)

torchaudio/functional/functional.py

@@ -1211,7 +1211,7 @@ def sliding_window_cmn(
     Apply sliding-window cepstral mean (and optionally variance) normalization per utterance.
 
     Args:
-        specgram (Tensor): Tensor of audio of dimension `(..., time, freq)`
+        specgram (Tensor): Tensor of spectrogram of dimension `(..., time, freq)`
         cmn_window (int, optional): Window in frames for running average CMN computation (int, default = 600)
         min_cmn_window (int, optional):  Minimum CMN window used at start of decoding (adds latency only at start).
             Only applicable if center == false, ignored if center==true (int, default = 100)

torchaudio/transforms.py

@@ -1198,17 +1198,17 @@ class SlidingWindowCmn(torch.nn.Module):
         self.center = center
         self.norm_vars = norm_vars
 
-    def forward(self, waveform: Tensor) -> Tensor:
+    def forward(self, specgram: Tensor) -> Tensor:
         r"""
         Args:
-            waveform (Tensor): Tensor of audio of dimension `(..., time)`.
+            specgram (Tensor): Tensor of spectrogram of dimension `(..., time, freq)`.
 
         Returns:
-            Tensor: Tensor of audio of dimension `(..., time)`.
+            Tensor: Tensor of spectrogram of dimension `(..., time, freq)`.
         """
-        cmn_waveform = F.sliding_window_cmn(
-            waveform, self.cmn_window, self.min_cmn_window, self.center, self.norm_vars)
-        return cmn_waveform
+        cmn_specgram = F.sliding_window_cmn(
+            specgram, self.cmn_window, self.min_cmn_window, self.center, self.norm_vars)
+        return cmn_specgram
 
 
 class Vad(torch.nn.Module):