《Python 语音转换简易速速上手小册》第8章 实时语音处理应用（2024 最新版）

pip install SpeechRecognition pyaudio

import speech_recognition as sr
def listen_and_recognize():
    # 初始化识别器和麦克风
    recognizer = sr.Recognizer()
    microphone = sr.Microphone()
    with microphone as source:
        print("请开始说话...")
        # 调整识别器的环境噪声水平
        recognizer.adjust_for_ambient_noise(source)
        audio = recognizer.listen(source)
    try:
        # 使用 Google Web Speech API 进行识别
        print("正在识别...")
        text = recognizer.recognize_google(audio)
        print(f"识别结果：{text}")
    except sr.UnknownValueError:
        print("无法理解音频")
    except sr.RequestError:
        print("Google Speech Recognition服务出错")
if __name__ == "__main__":
    listen_and_recognize()

pip install SpeechRecognition pyaudio

import speech_recognition as sr
def recognize_from_microphone(recognizer, microphone):
    with microphone as source:
        recognizer.adjust_for_ambient_noise(source)
        print("正在录音，请开始会议对话...")
        audio = recognizer.listen(source)
    try:
        print("正在转录...")
        return recognizer.recognize_google(audio)
    except sr.RequestError:
        return "API 请求失败"
    except sr.UnknownValueError:
        return "无法识别的语音"
def main():
    recognizer = sr.Recognizer()
    mics = [sr.Microphone(device_index=i) for i in range(2)]  # 假设有两个麦克风
    for mic in mics:
        text = recognize_from_microphone(recognizer, mic)
        print(f"转录结果：{text}")
if __name__ == "__main__":
    main()

pip install SpeechRecognition pyaudio googletrans==4.0.0-rc1

import speech_recognition as sr
from googletrans import Translator
def translate_and_recognize(speech_recognizer, microphone, source_language, target_language):
    with microphone as source:
        print(f"请以 {source_language} 说话...")
        audio = speech_recognizer.listen(source)
    try:
        recognized_text = speech_recognizer.recognize_google(audio, language=source_language)
        print(f"识别的 {source_language} 文本: {recognized_text}")
        translator = Translator()
        translated_text = translator.translate(recognized_text, src=source_language, dest=target_language).text
        print(f"翻译为 {target_language} 的文本: {translated_text}")
    except sr.UnknownValueError:
        print("无法识别语音")
    except sr.RequestError:
        print("Google Speech Recognition服务出错")
    except Exception as e:
        print(f"翻译时出现问题: {e}")
def main():
    speech_recognizer = sr.Recognizer()
    microphone = sr.Microphone()
    # 设置源语言和目标语言
    source_language = 'en'  # 英语
    target_language = 'es'  # 西班牙语
    translate_and_recognize(speech_recognizer, microphone, source_language, target_language)
if __name__ == "__main__":
    main()

pip install pyaudio

import pyaudio
import numpy as np
CHUNK = 1024  # 缓冲区大小
FORMAT = pyaudio.paInt16  # 采样位深
CHANNELS = 1  # 单声道
RATE = 44100  # 采样率
def sound_detected(data, threshold):
    """检测声音是否超过阈值"""
    amplitude = np.frombuffer(data, dtype=np.int16)
    if np.abs(amplitude).mean() > threshold:
        return True
    return False
def main():
    audio = pyaudio.PyAudio()
    stream = audio.open(format=FORMAT, channels=CHANNELS, 
                       rate=RATE, input=True, 
                       frames_per_buffer=CHUNK)
    print("开始实时音频监控...")
    try:
        while True:
            data = stream.read(CHUNK)
            if sound_detected(data, threshold=1000):
                print("检测到声音！")
    except KeyboardInterrupt:
        print("实时音频监控结束")
    finally:
        stream.stop_stream()
        stream.close()
        audio.terminate()
if __name__ == "__main__":
    main()

pip install pyaudio librosa

import pyaudio
import numpy as np
import librosa
CHUNK = 1024  # 缓冲区大小
FORMAT = pyaudio.paFloat32  # 采样位深
CHANNELS = 1  # 单声道
RATE = 22050  # 采样率
def detect_beat(data):
    """检测音频数据中的节奏"""
    y = np.frombuffer(data, dtype=np.float32)
    tempo, _ = librosa.beat.beat_track(y=y, sr=RATE)
    return tempo
def main():
    audio = pyaudio.PyAudio()
    stream = audio.open(format=FORMAT, channels=CHANNELS, 
                       rate=RATE, input=True, 
                       frames_per_buffer=CHUNK)
    print("开始实时音乐节奏检测...")
    try:
        while True:
            data = stream.read(CHUNK)
            tempo = detect_beat(data)
            print(f"检测到节奏: {tempo} BPM")
    except KeyboardInterrupt:
        print("实时音乐节奏检测结束")
    finally:
        stream.stop_stream()
        stream.close()
        audio.terminate()
if __name__ == "__main__":
    main()

pip install pyaudio scipy

import pyaudio
import numpy as np
from scipy.signal import lfilter
CHUNK = 1024  # 缓冲区大小
FORMAT = pyaudio.paInt16  # 采样位深
CHANNELS = 1  # 单声道
RATE = 44100  # 采样率
def noise_reduction(data, rate):
    """简单的降噪处理"""
    n = int(rate / 1000)
    reduced_noise = lfilter([1.0 / n] * n, 1, data)
    return reduced_noise
def main():
    audio = pyaudio.PyAudio()
    stream = audio.open(format=FORMAT, channels=CHANNELS, 
                       rate=RATE, input=True, 
                       frames_per_buffer=CHUNK)
    print("开始实时语音增强...")
    try:
        while True:
            data = np.frombuffer(stream.read(CHUNK), dtype=np.int16)
            enhanced_audio = noise_reduction(data, RATE)
            # 在此处可以进一步处理增强后的音频
    except KeyboardInterrupt:
        print("实时语音增强结束")
    finally:
        stream.stop_stream()
        stream.close()
        audio.terminate()
if __name__ == "__main__":
    main()

pip install pyaudio

import pyaudio
import socket
import threading
# 音频流参数
CHUNK = 1024
FORMAT = pyaudio.paInt16
CHANNELS = 1
RATE = 44100
# 网络参数
HOST = '127.0.0.1'  # 或服务器IP
PORT = 12345
def audio_stream(sock):
    p = pyaudio.PyAudio()
    stream = p.open(format=FORMAT, channels=CHANNELS, rate=RATE, input=True, frames_per_buffer=CHUNK)
    try:
        while True:
            data = stream.read(CHUNK)
            sock.sendall(data)
    finally:
        stream.stop_stream()
        stream.close()
        p.terminate()
def main():
    sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
    sock.connect((HOST, PORT))
    print("已连接到服务器，开始语音通信...")
    audio_thread = threading.Thread(target=audio_stream, args=(sock,))
    audio_thread.start()
    try:
        while True:
            data = sock.recv(CHUNK)
            # 这里可以添加音频播放逻辑
    except KeyboardInterrupt:
        print("实时语音通信结束")
    finally:
        sock.close()
if __name__ == "__main__":
    main()

pip install SpeechRecognition pyaudio

import speech_recognition as sr
import requests
def send_command_to_device(command):
    """根据语音命令控制智能设备"""
    url = "http://your-smart-home-api/command"
    data = {"command": command}
    response = requests.post(url, json=data)
    return response.status_code == 200
def recognize_commands(recognizer, microphone):
    with microphone as source:
        print("请说出控制命令...")
        audio = recognizer.listen(source)
    try:
        command = recognizer.recognize_google(audio)
        print(f"识别的命令: {command}")
        success = send_command_to_device(command)
        if success:
            print("命令执行成功！")
        else:
            print("命令执行失败")
    except sr.UnknownValueError:
        print("无法理解语音")
    except sr.RequestError:
        print("Google Speech Recognition服务出错")
def main():
    recognizer = sr.Recognizer()
    microphone = sr.Microphone()
    recognize_commands(recognizer, microphone)
if __name__ == "__main__":
    main()

pip install SpeechRecognition pyaudio googletrans==4.0.0-rc1

import speech_recognition as sr
from googletrans import Translator
def translate_speech(recognizer, microphone, src_language, dest_language):
    with microphone as source:
        print("请说话...")
        audio = recognizer.listen(source)
    try:
        # 语音识别
        text = recognizer.recognize_google(audio, language=src_language)
        print(f"识别的 {src_language} 文本: {text}")
        # 文本翻译
        translator = Translator()
        translation = translator.translate(text, src=src_language, dest=dest_language)
        print(f"翻译为 {dest_language} 的文本: {translation.text}")
    except sr.UnknownValueError:
        print("无法识别语音")
    except sr.RequestError:
        print("语音识别服务出错")
    except Exception as e:
        print(f"翻译时出现问题: {e}")
def main():
    recognizer = sr.Recognizer()
    microphone = sr.Microphone()
    translate_speech(recognizer, microphone, 'en', 'es')  # 从英语到西班牙语
if __name__ == "__main__":
    main()