Skip to content

Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.

Fix #2793: convert stereo/multi-channel audio to mono in extract_fbank (#2819)

Daniel Batesa1bc5b59967e
diff --git a/funasr/utils/load_utils.py b/funasr/utils/load_utils.py
index 8377979e0..501e817c5 100644
--- a/funasr/utils/load_utils.py
+++ b/funasr/utils/load_utils.py
@@ -233,10 +233,14 @@ def extract_fbank(data, data_len=None, data_type: str = "sound", frontend=None,
data = torch.from_numpy(data)
if len(data.shape) < 2:
data = data[None, :] # data: [batch, N]
+ elif data.shape[0] > 1:
+ data = data.mean(dim=0, keepdim=True) # convert stereo/multi-channel to mono
data_len = [data.shape[1]] if data_len is None else data_len
elif isinstance(data, torch.Tensor):
if len(data.shape) < 2:
data = data[None, :] # data: [batch, N]
+ elif data.shape[0] > 1:
+ data = data.mean(dim=0, keepdim=True) # convert stereo/multi-channel to mono
data_len = [data.shape[1]] if data_len is None else data_len
elif isinstance(data, (list, tuple)):
data_list, data_len = [], []