|
|
@@ -20,7 +20,7 @@ import os.path
|
|
|
from funasr.datasets.dataset import ESPnetDataset
|
|
|
|
|
|
|
|
|
-SUPPORT_AUDIO_TYPE_SETS = ['flac', 'mp3', 'm4a', 'ogg', 'opus', 'wav', 'wma']
|
|
|
+SUPPORT_AUDIO_TYPE_SETS = ['flac', 'mp3', 'ogg', 'opus', 'wav', 'pcm']
|
|
|
|
|
|
def load_kaldi(input):
|
|
|
retval = kaldiio.load_mat(input)
|
|
|
@@ -60,9 +60,14 @@ def load_bytes(input):
|
|
|
array = np.frombuffer((middle_data.astype(dtype) - offset) / abs_max, dtype=np.float32)
|
|
|
return array
|
|
|
|
|
|
+def load_pcm(input):
|
|
|
+ with open(input,"rb") as f:
|
|
|
+ bytes = f.read()
|
|
|
+ return load_bytes(bytes)
|
|
|
|
|
|
DATA_TYPES = {
|
|
|
"sound": lambda x: torchaudio.load(x)[0][0].numpy(),
|
|
|
+ "pcm": load_pcm,
|
|
|
"kaldi_ark": load_kaldi,
|
|
|
"bytes": load_bytes,
|
|
|
"waveform": lambda x: x,
|
|
|
@@ -219,6 +224,9 @@ class IterableESPnetDataset(IterableDataset):
|
|
|
if audio_type not in SUPPORT_AUDIO_TYPE_SETS:
|
|
|
raise NotImplementedError(
|
|
|
f'Not supported audio type: {audio_type}')
|
|
|
+ if audio_type == "pcm":
|
|
|
+ _type = "pcm"
|
|
|
+
|
|
|
func = DATA_TYPES[_type]
|
|
|
array = func(value)
|
|
|
if self.fs is not None and name == "speech":
|
|
|
@@ -318,6 +326,8 @@ class IterableESPnetDataset(IterableDataset):
|
|
|
if audio_type not in SUPPORT_AUDIO_TYPE_SETS:
|
|
|
raise NotImplementedError(
|
|
|
f'Not supported audio type: {audio_type}')
|
|
|
+ if audio_type == "pcm":
|
|
|
+ _type = "pcm"
|
|
|
func = DATA_TYPES[_type]
|
|
|
# Load entry
|
|
|
array = func(value)
|
hnluo