Python语音转文字代码故障求助：speech_recognition库问题排查

Hey there! Let's break down why your speech recognition code might not be working as expected. I’ll go through the most likely issues and fixes based on the snippet you shared.

First, here’s your code for reference:

import speech_recognition as sr
r = sr.Recognizer()
with sr.Microphone() as source:
    r.adjust_for_ambient_noise(source)
    while True:
        print("said")
        audio = r.listen(source)
        try:
            print("You said " + r.recognize_google(audio))
        except LookupError:
            print("i do not understand audio")

Now let’s look at potential problems:

• Microphone Permission Issues
  This is one of the most common culprits. If your Python script can’t access your microphone, the whole thing falls apart.

  • On Windows: Check your system settings to make sure Python has permission to use the microphone (Settings > Privacy & Security > Microphone).
  • On macOS: Go to System Preferences > Security & Privacy > Microphone, then make sure the terminal or IDE you’re using to run the script is checked.
  • On Linux: You might need to add your user to the audio group with sudo usermod -aG audio $USER, then log out and back in.

• Missing PyAudio Dependency
  The speech_recognition library relies on PyAudio to interface with your microphone. If you haven’t installed it, you’ll get an error when initializing the Microphone() object. Install it with:

  pip install pyaudio
  

  Note: On Windows, if you run into installation errors, you might need to download a precompiled wheel matching your Python version and architecture, then install it locally.

• Incomplete Exception Handling
  You’re only catching LookupError, but there are several other exceptions that can occur here, which would crash your script without giving you useful info. Expand your exception handling to catch these:

  try:
      print("You said " + r.recognize_google(audio))
  except sr.UnknownValueError:
      print("I couldn't understand what you said")
  except sr.RequestError as e:
      print(f"Couldn't connect to Google's service: {e}")
  except OSError:
      print("Failed to access the microphone")
  except LookupError:
      print("Unexpected error with speech recognition lookup")
  

  This way, you’ll get specific messages that tell you exactly what’s wrong—like no internet access for Google’s API, or a microphone hardware issue.

• Ambient Noise Calibration Problems
  The adjust_for_ambient_noise(source) call calibrates the recognizer to ignore background noise, but if it’s too quick (default is 0.5 seconds), it might not get an accurate reading. Try increasing the duration to give it more time to calibrate:

  r.adjust_for_ambient_noise(source, duration=1)
  

  This helps the recognizer distinguish your voice from background sounds better.

• Unbounded Loop (Quality of Life Issue)
  Your code runs in an infinite loop, which means you can only stop it by pressing Ctrl+C. Not a bug, but adding an exit condition makes it more user-friendly. For example:

  text = r.recognize_google(audio)
  print("You said " + text)
  if text.lower() == "exit":
      print("Exiting...")
      break
  

内容的提问来源于stack exchange，提问作者TanA