Home | 简体中文 | 繁体中文 | 杂文 | Github | 知乎专栏 | Facebook | Linkedin | Youtube | 打赏(Donations) | About
知乎专栏

15.5. Automatic Speech Recognition

15.5.1. kaldi

         
docker run -it kaldiasr/kaldi:latest bash
docker run -it --runtime=nvidia kaldiasr/kaldi:gpu-latest bash
        
		
         
docker run -it kaldiasr/kaldi:latest bash        
        
		

15.5.2. OpenAI Whisper

https://github.com/openai/whisper

         
import openai
audio_file= open("/path/to/file/audio.mp3", "rb")
transcript = openai.Audio.transcribe("whisper-1", audio_file)                
        
		

15.5.3. FunASR - 语音识别

15.5.3.1. 安装与演示

    	
pip install --upgrade pip    	
pip install -U funasr modelscope
    	
			
    	
from funasr import AutoModel
from funasr.utils.postprocess_utils import rich_transcription_postprocess

model_dir = "iic/SenseVoiceSmall"


model = AutoModel(
    model=model_dir,
    # trust_remote_code=True,
    # remote_code="./model.py",
    vad_model="fsmn-vad",
    vad_kwargs={"max_single_segment_time": 30000},
    # device="cuda:0",
    disable_update=True
)

# en
res = model.generate(
    input=f"{model.model_path}/example/zh.mp3",
    cache={},
    language="auto",  # "zn", "en", "yue", "ja", "ko", "nospeech"
    use_itn=True,
    batch_size_s=60,
    merge_vad=True,  #
    merge_length_s=15,
)
text = rich_transcription_postprocess(res[0]["text"])
print(text)
    	
			
    		
/Users/neo/tmp/social/.venv/bin/python /Users/neo/tmp/social/test.py 
funasr version: 1.3.0.
Downloading Model from https://www.modelscope.cn to directory: /Users/neo/.cache/modelscope/hub/models/iic/SenseVoiceSmall
WARNING:root:trust_remote_code: False
Downloading Model from https://www.modelscope.cn to directory: /Users/neo/.cache/modelscope/hub/models/iic/speech_fsmn_vad_zh-cn-16k-common-pytorch
WARNING:root:trust_remote_code: False
rtf_avg: 0.009: 100%|██████████| 1/1 [00:00<00:00, 19.43it/s]
  0%|          | 0/1 [00:00<?, ?it/s]
  0%|          | 0/1 [00:00<?, ?it/s]
100%|██████████| 1/1 [00:00<00:00,  1.86it/s]
{'load_data': '0.000', 'extract_feat': '0.003', 'forward': '0.538', 'batch_size': '1', 'rtf': '0.104'}, : 100%|██████████| 1/1 [00:00<00:00,  1.86it/s]
rtf_avg: 0.104: 100%|██████████| 1/1 [00:00<00:00,  1.86it/s]
rtf_avg: 0.096, time_speech:  5.616, time_escape: 0.540: 100%|██████████| 1/1 [00:00<00:00,  1.77it/s]
开放时间早上9点至下午5点。

Process finished with exit code 0