{"id":1176672,"date":"2025-01-15T17:47:47","date_gmt":"2025-01-15T09:47:47","guid":{"rendered":"https:\/\/docs.pingcode.com\/ask\/ask-ask\/1176672.html"},"modified":"2025-01-15T17:47:50","modified_gmt":"2025-01-15T09:47:50","slug":"%e5%a6%82%e4%bd%95%e7%94%a8python%e8%af%ad%e9%9f%b3%e8%bd%ac%e6%96%87%e5%ad%97","status":"publish","type":"post","link":"https:\/\/docs.pingcode.com\/ask\/1176672.html","title":{"rendered":"\u5982\u4f55\u7528python\u8bed\u97f3\u8f6c\u6587\u5b57"},"content":{"rendered":"

\"\u5982\u4f55\u7528python\u8bed\u97f3\u8f6c\u6587\u5b57\"<\/p>\n

\u4f7f\u7528Python\u8fdb\u884c\u8bed\u97f3\u8f6c\u6587\u5b57\u7684\u65b9\u6cd5\u6709\u591a\u79cd\uff0c\u5305\u62ec\u4f7f\u7528Google Speech Recognition API\u3001IBM Watson Speech to Text API\u3001Microsoft Azure Speech API\u3001\u4ee5\u53ca\u5f00\u6e90\u7684CMU Sphinx\u7b49\u3002<\/strong> \u5176\u4e2d\uff0cGoogle Speech Recognition API<\/strong> \u662f\u6700\u5e38\u7528\u548c\u7b80\u5355\u7684\u65b9\u6cd5\u4e4b\u4e00\u3002\u672c\u6587\u5c06\u8be6\u7ec6\u4ecb\u7ecd\u5982\u4f55\u4f7f\u7528\u8fd9\u4e9b\u5de5\u5177\u5b9e\u73b0\u8bed\u97f3\u8f6c\u6587\u5b57\u7684\u529f\u80fd\uff0c\u5e76\u63d0\u4f9b\u4ee3\u7801\u793a\u4f8b\u3002<\/p>\n<\/p>\n

\u4e00\u3001GOOGLE SPEECH RECOGNITION API<\/h3>\n<\/p>\n

Google Speech Recognition API \u662f\u4e00\u4e2a\u5f3a\u5927\u4e14\u6613\u4e8e\u4f7f\u7528\u7684\u5de5\u5177\u3002\u5b83\u652f\u6301\u591a\u79cd\u8bed\u8a00\uff0c\u5e76\u4e14\u80fd\u591f\u5904\u7406\u80cc\u666f\u566a\u97f3\u3002\u4f7f\u7528\u6b64API\u9700\u8981\u5b89\u88c5 speech_recognition<\/code> \u5e93\u3002<\/p>\n<\/p>\n

1. \u5b89\u88c5 SpeechRecognition \u5e93<\/h4>\n<\/p>\n

\u9996\u5148\uff0c\u5b89\u88c5 SpeechRecognition<\/code> \u5e93\u3002\u4f60\u53ef\u4ee5\u4f7f\u7528\u4ee5\u4e0b\u547d\u4ee4\u5728\u7ec8\u7aef\u6216\u547d\u4ee4\u63d0\u793a\u7b26\u4e2d\u8fdb\u884c\u5b89\u88c5\uff1a<\/p>\n<\/p>\n

pip install SpeechRecognition<\/p>\n
<\/code><\/pre>\n<\/p>\n
2. \u5f55\u5236\u97f3\u9891<\/h4>\n<\/p>\n
\u4f60\u53ef\u4ee5\u4f7f\u7528 pyaudio<\/code> \u5e93\u6765\u5f55\u5236\u97f3\u9891\u3002\u9996\u5148\uff0c\u5b89\u88c5 pyaudio<\/code>\uff1a<\/p>\n<\/p>\n
pip install pyaudio<\/p>\n
<\/code><\/pre>\n<\/p>\n
3. \u8bed\u97f3\u8f6c\u6587\u5b57\u4ee3\u7801\u793a\u4f8b<\/h4>\n<\/p>\n
\u4ee5\u4e0b\u662f\u4e00\u4e2a\u793a\u4f8b\u4ee3\u7801\uff0c\u5c55\u793a\u4e86\u5982\u4f55\u4f7f\u7528 speech_recognition<\/code> \u5e93\u5c06\u8bed\u97f3\u8f6c\u6362\u4e3a\u6587\u5b57\uff1a<\/p>\n<\/p>\n
import speech_recognition as sr<\/p>\n
def recognize_speech_from_mic(recognizer, microphone):<\/p>\n
    """Transcribe speech from recorded from `microphone`."""<\/p>\n
    # check that recognizer and microphone arguments are appropriate type<\/p>\n
    if not isinstance(recognizer, sr.Recognizer):<\/p>\n
        rAI<\/a>se TypeError("`recognizer` must be `Recognizer` instance")<\/p>\n
    if not isinstance(microphone, sr.Microphone):<\/p>\n
        raise TypeError("`microphone` must be `Microphone` instance")<\/p>\n
    # adjust the recognizer sensitivity to ambient noise and record audio from the microphone<\/p>\n
    with microphone as source:<\/p>\n
        recognizer.adjust_for_ambient_noise(source)<\/p>\n
        audio = recognizer.listen(source)<\/p>\n
    # set up the response object<\/p>\n
    response = {<\/p>\n
        "success": True,<\/p>\n
        "error": None,<\/p>\n
        "transcription": None<\/p>\n
    }<\/p>\n
    # try recognizing the speech in the recording<\/p>\n
    try:<\/p>\n
        response["transcription"] = recognizer.recognize_google(audio)<\/p>\n
    except sr.RequestError:<\/p>\n
        # API was unreachable or unresponsive<\/p>\n
        response["success"] = False<\/p>\n
        response["error"] = "API unavailable"<\/p>\n
    except sr.UnknownValueError:<\/p>\n
        # speech was unintelligible<\/p>\n
        response["error"] = "Unable to recognize speech"<\/p>\n
    return response<\/p>\n
if __name__ == "__main__":<\/p>\n
    recognizer = sr.Recognizer()<\/p>\n
    microphone = sr.Microphone()<\/p>\n
    print("Please speak something...")<\/p>\n
    response = recognize_speech_from_mic(recognizer, microphone)<\/p>\n
    if response["success"]:<\/p>\n
        print("You said: {}".format(response["transcription"]))<\/p>\n
    else:<\/p>\n
        print("I didn't catch that. What did you say?\\nError: {}".format(response["error"]))<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u4e8c\u3001IBM WATSON SPEECH TO TEXT API<\/h3>\n<\/p>\n
IBM Watson Speech to Text API \u662f\u53e6\u4e00\u4e2a\u529f\u80fd\u5f3a\u5927\u7684\u8bed\u97f3\u8bc6\u522b\u5de5\u5177\uff0c\u652f\u6301\u591a\u79cd\u8bed\u8a00\u548c\u65b9\u8a00\uff0c\u5e76\u4e14\u63d0\u4f9b\u4e86\u9ad8\u7ea7\u7684\u81ea\u5b9a\u4e49\u9009\u9879\u3002<\/p>\n<\/p>\n
1. \u83b7\u53d6 API \u5bc6\u94a5<\/h4>\n<\/p>\n
\u9996\u5148\uff0c\u4f60\u9700\u8981\u6ce8\u518c IBM Cloud \u5e76\u83b7\u53d6 API \u5bc6\u94a5\u3002\u6ce8\u518c\u5b8c\u6210\u540e\uff0c\u4f60\u53ef\u4ee5\u5728 IBM Cloud \u4eea\u8868\u677f\u4e2d\u521b\u5efa\u4e00\u4e2a Speech to Text \u670d\u52a1\u5b9e\u4f8b\uff0c\u5e76\u83b7\u53d6 API \u5bc6\u94a5\u548c\u670d\u52a1 URL\u3002<\/p>\n<\/p>\n
2. \u5b89\u88c5 IBM Watson SDK<\/h4>\n<\/p>\n
\u5b89\u88c5 ibm_watson<\/code> \u5e93\uff1a<\/p>\n<\/p>\n
pip install ibm_watson<\/p>\n
<\/code><\/pre>\n<\/p>\n
3. \u8bed\u97f3\u8f6c\u6587\u5b57\u4ee3\u7801\u793a\u4f8b<\/h4>\n<\/p>\n
\u4ee5\u4e0b\u662f\u4e00\u4e2a\u793a\u4f8b\u4ee3\u7801\uff0c\u5c55\u793a\u4e86\u5982\u4f55\u4f7f\u7528 IBM Watson Speech to Text API \u5c06\u8bed\u97f3\u8f6c\u6362\u4e3a\u6587\u5b57\uff1a<\/p>\n<\/p>\n
import json<\/p>\n
from ibm_watson import SpeechToTextV1<\/p>\n
from ibm_cloud_sdk_core.authenticators import IAMAuthenticator<\/p>\n
Replace 'YOUR_API_KEY' and 'YOUR_SERVICE_URL' with your actual API key and service URL<\/strong><\/h2>\n
api_key = 'YOUR_API_KEY'<\/p>\n
service_url = 'YOUR_SERVICE_URL'<\/p>\n
authenticator = IAMAuthenticator(api_key)<\/p>\n
speech_to_text = SpeechToTextV1(authenticator=authenticator)<\/p>\n
speech_to_text.set_service_url(service_url)<\/p>\n
with open('audio-file.wav', 'rb') as audio_file:<\/p>\n
    result = speech_to_text.recognize(<\/p>\n
        audio=audio_file,<\/p>\n
        content_type='audio\/wav',<\/p>\n
    ).get_result()<\/p>\n
print(json.dumps(result, indent=2))<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u4e09\u3001MICROSOFT AZURE SPEECH API<\/h3>\n<\/p>\n
Microsoft Azure Speech API \u662f\u4e00\u4e2a\u5f3a\u5927\u7684\u8bed\u97f3\u8bc6\u522b\u5de5\u5177\uff0c\u652f\u6301\u591a\u79cd\u8bed\u8a00\uff0c\u5e76\u4e14\u63d0\u4f9b\u4e86\u9ad8\u7ea7\u529f\u80fd\uff0c\u5982\u8bed\u97f3\u5408\u6210\u548c\u8bed\u97f3\u8bc6\u522b\u81ea\u5b9a\u4e49\u3002<\/p>\n<\/p>\n
1. \u83b7\u53d6 API \u5bc6\u94a5<\/h4>\n<\/p>\n
\u9996\u5148\uff0c\u4f60\u9700\u8981\u6ce8\u518c Microsoft Azure \u5e76\u83b7\u53d6 API \u5bc6\u94a5\u3002\u6ce8\u518c\u5b8c\u6210\u540e\uff0c\u4f60\u53ef\u4ee5\u5728 Azure Portal \u4e2d\u521b\u5efa\u4e00\u4e2a Speech \u670d\u52a1\u5b9e\u4f8b\uff0c\u5e76\u83b7\u53d6 API \u5bc6\u94a5\u548c\u670d\u52a1\u533a\u57df\u3002<\/p>\n<\/p>\n
2. \u5b89\u88c5 Azure SDK<\/h4>\n<\/p>\n
\u5b89\u88c5 azure-cognitiveservices-speech<\/code> \u5e93\uff1a<\/p>\n<\/p>\n
pip install azure-cognitiveservices-speech<\/p>\n
<\/code><\/pre>\n<\/p>\n
3. \u8bed\u97f3\u8f6c\u6587\u5b57\u4ee3\u7801\u793a\u4f8b<\/h4>\n<\/p>\n
\u4ee5\u4e0b\u662f\u4e00\u4e2a\u793a\u4f8b\u4ee3\u7801\uff0c\u5c55\u793a\u4e86\u5982\u4f55\u4f7f\u7528 Microsoft Azure Speech API \u5c06\u8bed\u97f3\u8f6c\u6362\u4e3a\u6587\u5b57\uff1a<\/p>\n<\/p>\n
import azure.cognitiveservices.speech as speechsdk<\/p>\n
Replace 'YOUR_SUBSCRIPTION_KEY' and 'YOUR_SERVICE_REGION' with your actual subscription key and service region<\/strong><\/h2>\n
subscription_key = 'YOUR_SUBSCRIPTION_KEY'<\/p>\n
service_region = 'YOUR_SERVICE_REGION'<\/p>\n
speech_config = speechsdk.SpeechConfig(subscription=subscription_key, region=service_region)<\/p>\n
audio_input = speechsdk.AudioConfig(filename="audio-file.wav")<\/p>\n
speech_recognizer = speechsdk.SpeechRecognizer(speech_config=speech_config, audio_config=audio_input)<\/p>\n
result = speech_recognizer.recognize_once()<\/p>\n
if result.reason == speechsdk.ResultReason.RecognizedSpeech:<\/p>\n
    print("Recognized: {}".format(result.text))<\/p>\n
elif result.reason == speechsdk.ResultReason.NoMatch:<\/p>\n
    print("No speech could be recognized")<\/p>\n
elif result.reason == speechsdk.ResultReason.Canceled:<\/p>\n
    cancellation_details = result.cancellation_details<\/p>\n
    print("Speech Recognition canceled: {}".format(cancellation_details.reason))<\/p>\n
    if cancellation_details.reason == speechsdk.CancellationReason.Error:<\/p>\n
        print("Error details: {}".format(cancellation_details.error_details))<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u56db\u3001CMU SPHINX<\/h3>\n<\/p>\n
CMU Sphinx \u662f\u4e00\u4e2a\u5f00\u6e90\u7684\u8bed\u97f3\u8bc6\u522b\u7cfb\u7edf\uff0c\u9002\u7528\u4e8e\u4e0d\u5e0c\u671b\u4f9d\u8d56\u7b2c\u4e09\u65b9API\u7684\u5e94\u7528\u7a0b\u5e8f\u3002<\/p>\n<\/p>\n
1. \u5b89\u88c5 pocketsphinx \u5e93<\/h4>\n<\/p>\n
\u5b89\u88c5 pocketsphinx<\/code> \u5e93\uff1a<\/p>\n<\/p>\n
pip install pocketsphinx<\/p>\n
<\/code><\/pre>\n<\/p>\n
2. \u8bed\u97f3\u8f6c\u6587\u5b57\u4ee3\u7801\u793a\u4f8b<\/h4>\n<\/p>\n
\u4ee5\u4e0b\u662f\u4e00\u4e2a\u793a\u4f8b\u4ee3\u7801\uff0c\u5c55\u793a\u4e86\u5982\u4f55\u4f7f\u7528 CMU Sphinx \u5c06\u8bed\u97f3\u8f6c\u6362\u4e3a\u6587\u5b57\uff1a<\/p>\n<\/p>\n
import speech_recognition as sr<\/p>\n
def recognize_speech_from_audio(audio_file):<\/p>\n
    recognizer = sr.Recognizer()<\/p>\n
    with sr.AudioFile(audio_file) as source:<\/p>\n
        audio = recognizer.record(source)<\/p>\n
    try:<\/p>\n
        text = recognizer.recognize_sphinx(audio)<\/p>\n
        print("You said: " + text)<\/p>\n
    except sr.UnknownValueError:<\/p>\n
        print("Sphinx could not understand audio")<\/p>\n
    except sr.RequestError as e:<\/p>\n
        print("Sphinx error; {0}".format(e))<\/p>\n
if __name__ == "__main__":<\/p>\n
    recognize_speech_from_audio('audio-file.wav')<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u4e94\u3001\u603b\u7ed3<\/h3>\n<\/p>\n
\u5728\u4f7f\u7528Python\u8fdb\u884c\u8bed\u97f3\u8f6c\u6587\u5b57\u65f6\uff0c\u53ef\u4ee5\u9009\u62e9\u4e0d\u540c\u7684API\u548c\u5de5\u5177\u6765\u5b9e\u73b0\uff0c\u6bcf\u79cd\u5de5\u5177\u90fd\u6709\u5176\u4f18\u52bf\u548c\u9002\u7528\u573a\u666f\uff1a<\/p>\n<\/p>\n
    \n
  • Google Speech Recognition API<\/strong>\uff1a\u6613\u4e8e\u4f7f\u7528\uff0c\u652f\u6301\u591a\u79cd\u8bed\u8a00\uff0c\u9002\u5408\u521d\u5b66\u8005\u3002<\/li>\n
  • IBM Watson Speech to Text API<\/strong>\uff1a\u63d0\u4f9b\u9ad8\u7ea7\u81ea\u5b9a\u4e49\u9009\u9879\uff0c\u9002\u5408\u9700\u8981\u9ad8\u7ea7\u529f\u80fd\u7684\u5e94\u7528\u3002<\/li>\n
  • Microsoft Azure Speech API<\/strong>\uff1a\u529f\u80fd\u5f3a\u5927\uff0c\u652f\u6301\u591a\u79cd\u8bed\u8a00\u548c\u9ad8\u7ea7\u529f\u80fd\uff0c\u9002\u5408\u4f01\u4e1a\u7ea7\u5e94\u7528\u3002<\/li>\n
  • CMU Sphinx<\/strong>\uff1a\u5f00\u6e90\uff0c\u4e0d\u4f9d\u8d56\u7b2c\u4e09\u65b9API\uff0c\u9002\u5408\u5bf9\u9690\u79c1\u548c\u6570\u636e\u5b89\u5168\u6709\u8f83\u9ad8\u8981\u6c42\u7684\u5e94\u7528\u3002<\/li>\n<\/ul>\n
    \u6839\u636e\u5177\u4f53\u9700\u6c42\u9009\u62e9\u5408\u9002\u7684\u5de5\u5177\uff0c\u5e76\u7ed3\u5408\u4ee3\u7801\u793a\u4f8b\u8fdb\u884c\u5b9e\u73b0\uff0c\u53ef\u4ee5\u5e2e\u52a9\u4f60\u5feb\u901f\u5b8c\u6210\u8bed\u97f3\u8f6c\u6587\u5b57\u7684\u529f\u80fd\u5f00\u53d1\u3002<\/p>\n<\/p>\n
    \u76f8\u5173\u95ee\u7b54FAQs\uff1a<\/strong><\/h2>\n
     \u5982\u4f55\u4f7f\u7528Python\u8fdb\u884c\u8bed\u97f3\u8bc6\u522b\uff1f<\/strong>
    Python\u63d0\u4f9b\u4e86\u591a\u79cd\u5e93\u6765\u5b9e\u73b0\u8bed\u97f3\u8f6c\u6587\u5b57\u7684\u529f\u80fd\u3002\u6700\u5e38\u7528\u7684\u5e93\u4e4b\u4e00\u662fSpeechRecognition<\/code>\uff0c\u5b83\u652f\u6301\u591a\u79cd\u8bed\u97f3\u8bc6\u522b\u5f15\u64ce\u548cAPI\uff0c\u5305\u62ecGoogle Web Speech API\u3001Sphinx\u7b49\u3002\u7528\u6237\u53ea\u9700\u5b89\u88c5\u76f8\u5173\u5e93\uff0c\u7f16\u5199\u7b80\u5355\u7684\u4ee3\u7801\uff0c\u5c31\u80fd\u8f7b\u677e\u5c06\u8bed\u97f3\u8f6c\u6362\u4e3a\u6587\u672c\u3002<\/p>\n
    \u6211\u9700\u8981\u54ea\u4e9b\u5de5\u5177\u6216\u5e93\u6765\u5b9e\u73b0\u8bed\u97f3\u8f6c\u6587\u5b57\uff1f<\/strong>
    \u8981\u4f7f\u7528Python\u8fdb\u884c\u8bed\u97f3\u8f6c\u6587\u5b57\uff0c\u60a8\u9700\u8981\u5b89\u88c5SpeechRecognition<\/code>\u5e93\u3002\u6b64\u5916\uff0cPyAudio<\/code>\u5e93\u7528\u4e8e\u5904\u7406\u97f3\u9891\u8f93\u5165\u4e5f\u975e\u5e38\u91cd\u8981\u3002\u5728\u7ec8\u7aef\u4e2d\u4f7f\u7528pip install SpeechRecognition pyaudio<\/code>\u547d\u4ee4\u5373\u53ef\u8f7b\u677e\u5b89\u88c5\u8fd9\u4e9b\u5e93\u3002<\/p>\n
    \u5982\u4f55\u63d0\u9ad8\u8bed\u97f3\u8bc6\u522b\u7684\u51c6\u786e\u6027\uff1f<\/strong>
    \u63d0\u9ad8\u8bed\u97f3\u8bc6\u522b\u51c6\u786e\u6027\u7684\u529e\u6cd5\u6709\u5f88\u591a\u3002\u9996\u5148\uff0c\u786e\u4fdd\u5f55\u97f3\u73af\u5883\u5b89\u9759\uff0c\u51cf\u5c11\u80cc\u666f\u566a\u97f3\u3002\u6b64\u5916\uff0c\u4f7f\u7528\u9ad8\u8d28\u91cf\u7684\u9ea6\u514b\u98ce\u53ef\u4ee5\u6709\u6548\u63d0\u5347\u97f3\u9891\u7684\u6e05\u6670\u5ea6\u3002\u540c\u65f6\uff0c\u53ef\u4ee5\u901a\u8fc7\u8c03\u6574\u8bc6\u522b\u5f15\u64ce\u7684\u53c2\u6570\uff0c\u6216\u5bf9\u7279\u5b9a\u9886\u57df\u7684\u672f\u8bed\u8fdb\u884c\u8bad\u7ec3\u6765\u8fdb\u4e00\u6b65\u63d0\u9ad8\u51c6\u786e\u6027\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"\u4f7f\u7528Python\u8fdb\u884c\u8bed\u97f3\u8f6c\u6587\u5b57\u7684\u65b9\u6cd5\u6709\u591a\u79cd\uff0c\u5305\u62ec\u4f7f\u7528Google Speech Recognition API […]","protected":false},"author":3,"featured_media":1176678,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[37],"tags":[],"acf":[],"_links":{"self":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/1176672"}],"collection":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/comments?post=1176672"}],"version-history":[{"count":"1","href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/1176672\/revisions"}],"predecessor-version":[{"id":1176679,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/1176672\/revisions\/1176679"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media\/1176678"}],"wp:attachment":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media?parent=1176672"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/categories?post=1176672"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/tags?post=1176672"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}