FakeVD/code_test/Text_Feature_Extractor/wav2text.py

from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks

# ./FakeVD/Models/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch/
# damo/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch
class wav2text:
    def __init__(self) -> None:
        self.model = pipeline('auto-speech-recognition',
                    './FakeVD/Models/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch/', device="cpu")

    def work(self, wav_path):
        return self.model(wav_path,)
    

if __name__ == "__main__":
    model = wav2text()
    result = model.work('./FakeVD/dataset/videos_1/douyin_6671891732524829965.mp4')
    print(result)