Python SpeechRecognition库无法获取音频，程序卡在Listening状态如何解决？

我尝试基于SpeechRecognition 3.8.1库实现语音采集功能已4天，先后查阅了GitHub对应仓库issue、GeeksforGeeks语音助手教程、Stack Overflow多个同类问题帖等大量资料，所有方案均无效。

最初执行sudo apt-get install python-pyaudio python3-pyaudio安装PyAudio失败，经过多次尝试后通过pipwin install pyaudio完成安装。但运行下述代码时，程序始终卡在输出Listening...的环节，无后续响应：

import os
import pyttsx3, datetime, pyaudio
import speech_recognition as sr

# Initial Setup for pyttsx3 - speaking abilities
engine = pyttsx3.init("sapi5")
voices = engine.getProperty("voices")
engine.setProperty("voice", voices[1].id)  # 0-male voice , 1-female voice
sr.Microphone.list_microphone_names()

# Initial Setup for speech_recognition - listening abilities
# r.energy_threshold = 10
# print(pyaudio.get_device_count() - 1)


def speak(speakable):
    """speak() takes a string and reads it loud"""
    engine.say(str(speakable))
    engine.runAndWait()


def takeCommand():
    pyaudio.PyAudio()
    r = sr.Recognizer()
    """It takes microphone input from the user and returns string output"""
    with sr.Microphone() as source:
        r.adjust_for_ambient_noise(source, duration=0.9)
        print("Listening...")
        r.pause_threshold = 45
        audio = ""
        try:
            audio = r.listen(source)
            print("Recognizing...")
        except Exception as e:
            print("Listen err: ", e)
        try:
            print("Recognizing...")
            query = r.recognize_google(audio)
            print(f"User said: {query}\n")  # User query will be printed.
        except sr.UnknownValueError as e:
            print("Say that again please...")
            return "None"  # None string will be returned
        except Exception as err:
            print("Check your internet...")
            return "None"
    return query


def wishMe():
    hour = int(datetime.datetime.now().hour)
    if hour >= 0 and hour < 12:
        speak("Good Morning!")

    elif hour >= 12 and hour < 18:
        speak("Good Afternoon!")

    else:
        speak("Good Evening!")
        speak(
            "Hello Sir, I am Friday, your Artificial intelligence assistant. Please tell me how may I help you"
        )


if __name__ == "__main__":
    os.system("CLS")
    while True:
        command = takeCommand().lower()
        print(f"Command: {command}")
        if "wish" in command:
            wishMe()

环境信息 #

• 运行环境为Windows 10 Home
• 代码编辑器为VS Code
• 本项目未使用virtual env虚拟环境
• 已验证Chrome语音搜索功能可正常使用，麦克风权限无异常
• 执行python3 -m speech_recognition命令的结果见下图：
  

按顺序修改以下代码问题即可解决卡顿问题：

1. 调整pause_threshold参数
   你设置的r.pause_threshold = 45代表语音结束后需要等待45秒才会进入识别环节，这是程序看起来卡住的核心原因，将该值调整为1即可，代表语音停顿1秒就判定为输入结束。
2. 删除多余的pyaudio初始化代码
   每次调用takeCommand()时都执行pyaudio.PyAudio()会重复创建实例，占用麦克风资源，直接删掉这行代码即可，speech_recognition库内部会自动管理pyaudio实例。
3. 指定麦克风设备索引
   系统存在多个音频输入设备时，默认选中的设备可能不是你正在使用的麦克风。运行print(sr.Microphone.list_microphone_names())输出所有设备列表，找到你在用的麦克风对应的索引值，修改sr.Microphone()为sr.Microphone(device_index=你查到的索引值)即可。
4. 修复audio初始值问题
   你将audio初始值设为空字符串，如果listen()环节报错，后续识别步骤传入空字符串会触发未知错误，将初始值改为None，识别前增加判空逻辑即可。

修改后的核心takeCommand函数参考：

def takeCommand():
    r = sr.Recognizer()
    # 替换device_index为你查到的麦克风索引
    with sr.Microphone(device_index=0) as source:
        r.adjust_for_ambient_noise(source, duration=0.9)
        print("Listening...")
        r.pause_threshold = 1
        audio = None
        try:
            # 增加超时限制避免程序无限等待
            audio = r.listen(source, timeout=10, phrase_time_limit=15)
            print("Recognizing...")
        except Exception as e:
            print("Listen err: ", e)
            return "None"
        try:
            # 如需识别中文可以加language='zh-CN'参数，不需要可删除
            query = r.recognize_google(audio)
            print(f"User said: {query}\n")
        except sr.UnknownValueError as e:
            print("Say that again please...")
            return "None"
        except Exception as err:
            print("Check your internet...")
            return "None"
    return query

内容的提问来源于stack exchange，提问作者Curious Learner