提交

kongdeqiang

2026-03-13 28ccfbfc51068a663a80764e14074df5edf2b5ba

 runtime/python/onnxruntime/demo_vad_online.py

@@ -3,7 +3,9 @@
from pathlib import Path

model_dir = "damo/speech_fsmn_vad_zh-cn-16k-common-pytorch"
wav_path = '{}/.cache/modelscope/hub/damo/speech_fsmn_vad_zh-cn-16k-common-pytorch/example/vad_example.wav'.format(Path.home())
wav_path = "{}/.cache/modelscope/hub/damo/speech_fsmn_vad_zh-cn-16k-common-pytorch/example/vad_example.wav".format(
    Path.home()
)

model = Fsmn_vad_online(model_dir)

@@ -14,15 +16,16 @@
#
sample_offset = 0
step = 1600
param_dict = {'in_cache': []}
param_dict = {"in_cache": []}
for sample_offset in range(0, speech_length, min(step, speech_length - sample_offset)):
    if sample_offset + step >= speech_length - 1:
        step = speech_length - sample_offset
        is_final = True
    else:
        is_final = False
    param_dict['is_final'] = is_final
    segments_result = model(audio_in=speech[sample_offset: sample_offset + step],
                            param_dict=param_dict)
    param_dict["is_final"] = is_final
    segments_result = model(
        audio_in=speech[sample_offset : sample_offset + step], param_dict=param_dict
    )
    if segments_result:
        print(segments_result)
        print(segments_result)