IBM waston speach to text

waston.py

from ibm_watson import SpeechToTextV1

speech_to_text = SpeechToTextV1(
    iam_apikey='{apikey}'.format(apikey="APIKEY"),
    url='{url}'.format(url="https://stream.watsonplatform.net/speech-to-text/api")
)

with open('audio-file.flac', 'rb')as audio_file:
    print(speech_to_text.recognize(
        audio_file, content_type='audio/flac', timestamps=True, model='zh-CN_NarrowbandModel', word_confidence=True
    ))

IBM云提供的语音转文本服务。 之前一个想法是给podcast做字幕用。