1） 生成需要识别的wav文件，SpeechRecognition需要wav文件，不能识别mp3文件

安装库：

sudo apt install espeak ffmpeg libespeak1
pip install pyttsx3

代码：

def demo_tts_wav():
    import pyttsx3

    engine = pyttsx3.init()

    engine.setProperty('rate', 150)
    engine.setProperty('volume', 1.0)
    voices = engine.getProperty('voices')
    engine.setProperty('voice', voices[0].id)

    text = '你好，我是一个AI机器人'
    #engine.say(text)
    filename = 'ni_hao.wav'
    engine.save_to_file(text, filename)

    engine.runAndWait()
 

2. 语音识别，使用speech_recognition

安装库：

pip install SpeechRecognition

pip install pocketsphinx

下载模型文件：CMU Sphinx - Browse /Acoustic and Language Models/Mandarin at SourceForge.net

pip install vosk

下载模型文件到代码目录下：VOSK Models

解压，并且重命名为model

代码

def demo_speech_recognition():
    import speech_recognition as sr

    r = sr.Recognizer()
    try:
        audio_file = sr.AudioFile('ni_hao.wav')
        with audio_file as source:
            audio_data = r.record(source)
        #text = r.recognize_google(audio_data, language='zh-Cn')
        #text = r.recognize_wit(audio_data)
        text = r.recognize_vosk(audio_data, language='zh-Cn')
        print("识别结果：", text)
    except Exception as e:
        print("无法识别语音：", str(e))