《Python 语音转换简易速速上手小册》第9章 特定领域的语音处理（2024 最新版）

pip install SpeechRecognition pyaudio nltk

import speech_recognition as sr
import nltk
from nltk.tokenize import word_tokenize
nltk.download('punkt')
def recognize_speech_from_mic(recognizer, microphone):
    with microphone as source:
        print("请问您需要什么帮助？")
        audio = recognizer.listen(source)
    try:
        return recognizer.recognize_google(audio)
    except sr.UnknownValueError:
        return "无法理解语音"
    except sr.RequestError:
        return "服务出错"
def respond_to_request(text):
    tokens = word_tokenize(text)
    # 根据关键词进行简单的自然语言处理
    if '退货' in tokens:
        return "关于退货，您可以..."
    elif '账单' in tokens or '支付' in tokens:
        return "关于账单和支付，您可以..."
    else:
        return "抱歉，我没有找到相关信息。"
def main():
    recognizer = sr.Recognizer()
    microphone = sr.Microphone()
    speech_text = recognize_speech_from_mic(recognizer, microphone)
    print(f"您说: {speech_text}")
    response = respond_to_request(speech_text)
    print(response)
if __name__ == "__main__":
    main()

pip install SpeechRecognition pyaudio

import speech_recognition as sr
def listen_to_doctor(recognizer, microphone):
    with microphone as source:
        print("请医生开始说话...")
        audio = recognizer.listen(source)
    try:
        return recognizer.recognize_google(audio)
    except sr.UnknownValueError:
        return "无法理解语音"
    except sr.RequestError:
        return "服务出错"
def format_medical_record(speech_text):
    # 这里可以添加更复杂的文本处理逻辑
    return f"病历记录: {speech_text}"
def main():
    recognizer = sr.Recognizer()
    microphone = sr.Microphone()
    speech_text = listen_to_doctor(recognizer, microphone)
    print(f"识别的语音: {speech_text}")
    formatted_record = format_medical_record(speech_text)
    print(formatted_record)
if __name__ == "__main__":
    main()

pip install SpeechRecognition pyaudio

import speech_recognition as sr
def recognize_speech_and_evaluate(recognizer, microphone, target_phrase):
    with microphone as source:
        print("请重复以下短语来练习发音：")
        print(target_phrase)
        audio = recognizer.listen(source)
    try:
        recognized_text = recognizer.recognize_google(audio)
        print(f"您的发音: {recognized_text}")
        # 简单的评估逻辑：比较目标短语和识别结果
        if recognized_text.lower() == target_phrase.lower():
            return "发音准确！"
        else:
            return "发音需要改进。"
    except sr.UnknownValueError:
        return "无法理解语音"
    except sr.RequestError:
        return "服务出错"
def main():
    recognizer = sr.Recognizer()
    microphone = sr.Microphone()
    target_phrase = "Good morning, how are you?"
    feedback = recognize_speech_and_evaluate(recognizer, microphone, target_phrase)
    print(feedback)
if __name__ == "__main__":
    main()

pip install SpeechRecognition pyaudio

import speech_recognition as sr
# 假设有一个预先定义的医疗术语词汇库
medical_terms = ['anemia', 'diabetes', 'hypertension', 'leukemia']
def recognize_medical_speech(recognizer, microphone):
    with microphone as source:
        print("请医生开始描述病情...")
        audio = recognizer.listen(source)
    try:
        recognized_text = recognizer.recognize_google(audio)
        print(f"识别的文本: {recognized_text}")
        return recognized_text
    except sr.UnknownValueError:
        return "无法理解语音"
    except sr.RequestError:
        return "服务出错"
def check_medical_terms(text, terms):
    """检查识别的文本中是否包含医疗术语"""
    return [term for term in terms if term in text.lower()]
def main():
    recognizer = sr.Recognizer()
    microphone = sr.Microphone()
    speech_text = recognize_medical_speech(recognizer, microphone)
    medical_terms_in_speech = check_medical_terms(speech_text, medical_terms)
    if medical_terms_in_speech:
        print("识别的医疗术语: " + ", ".join(medical_terms_in_speech))
    else:
        print("未识别到医疗术语。")
if __name__ == "__main__":
    main()

pip install SpeechRecognition pyaudio

import speech_recognition as sr
# 假设我们有不同方言的语音数据
dialects = {
    "mandarin": "zh-CN",  # 普通话
    "cantonese": "yue-Hant-HK",  # 粤语
    "shanghainese": "wuu-CN"  # 上海话（注：实际应用中可能需要特定支持）
}
def recognize_speech_dialect(recognizer, microphone, dialect):
    with microphone as source:
        print(f"请以{dialect}方言说话...")
        audio = recognizer.listen(source)
    try:
        language = dialects.get(dialect, "zh-CN")
        recognized_text = recognizer.recognize_google(audio, language=language)
        print(f"识别的文本: {recognized_text}")
        return recognized_text
    except sr.UnknownValueError:
        return "无法理解语音"
    except sr.RequestError:
        return "服务出错"
def main():
    recognizer = sr.Recognizer()
    microphone = sr.Microphone()
    # 测试不同的方言
    for dialect in dialects:
        recognize_speech_dialect(recognizer, microphone, dialect)
if __name__ == "__main__":
    main()

pip install SpeechRecognition pyaudio noisereduce

import speech_recognition as sr
import noisereduce as nr
import numpy as np
def recognize_speech_in_noise(recognizer, microphone):
    with microphone as source:
        recognizer.adjust_for_ambient_noise(source)
        print("请说话...")
        audio = recognizer.listen(source)
    # 将音频数据转换为NumPy数组
    audio_data = np.frombuffer(audio.get_wav_data(), dtype=np.int16)
    # 应用噪声减少
    reduced_noise_audio = nr.reduce_noise(y=audio_data, sr=16000)
    try:
        # 识别处理过的音频
        recognized_text = recognizer.recognize_google(audio_data=reduced_noise_audio, language="zh-CN")
        print(f"识别的文本: {recognized_text}")
        return recognized_text
    except sr.UnknownValueError:
        return "无法理解语音"
    except sr.RequestError:
        return "服务出错"
def main():
    recognizer = sr.Recognizer()
    microphone = sr.Microphone()
    recognize_speech_in_noise(recognizer, microphone)
if __name__ == "__main__":
    main()

pip install SpeechRecognition pyaudio

import speech_recognition as sr
def transcribe_medical_record(recognizer, microphone):
    with microphone as source:
        print("请医生开始说话，进行病历记录...")
        audio = recognizer.listen(source)
    try:
        # 使用 Google 语音识别进行转录
        transcript = recognizer.recognize_google(audio, language='zh-CN')
        return transcript
    except sr.UnknownValueError:
        return "无法理解语音"
    except sr.RequestError:
        return "语音识别服务出错"
def format_transcript(transcript):
    # 对转录的文本进行格式化处理，以适应病历记录的标准
    # 这里可以根据需要添加更多的文本处理逻辑
    formatted_transcript = "病历记录:\n" + transcript
    return formatted_transcript
def main():
    recognizer = sr.Recognizer()
    microphone = sr.Microphone()
    transcript = transcribe_medical_record(recognizer, microphone)
    print("转录内容：", transcript)
    formatted_transcript = format_transcript(transcript)
    print("格式化后的病历记录：\n", formatted_transcript)
if __name__ == "__main__":
    main()

pip install SpeechRecognition pyaudio

import speech_recognition as sr
def recognize_trade_command(recognizer, microphone):
    with microphone as source:
        print("请说出您的交易指令...")
        audio = recognizer.listen(source)
    try:
        command = recognizer.recognize_google(audio)
        return command
    except sr.UnknownValueError:
        return "无法理解语音"
    except sr.RequestError:
        return "服务出错"
def process_trade_command(command):
    # 这里可以添加解析命令的逻辑，如买入、卖出、查询等
    if "买入" in command:
        return "执行买入操作"
    elif "卖出" in command:
        return "执行卖出操作"
    else:
        return "未识别的交易指令"
def main():
    recognizer = sr.Recognizer()
    microphone = sr.Microphone()
    command = recognize_trade_command(recognizer, microphone)
    response = process_trade_command(command)
    print("交易助手回应：", response)
if __name__ == "__main__":
    main()

pip install SpeechRecognition pyaudio googletrans==4.0.0-rc1

import speech_recognition as sr
from googletrans import Translator
def recognize_customer_speech(recognizer, microphone, language="en"):
    with microphone as source:
        print(f"请用 {language} 说话...")
        audio = recognizer.listen(source)
    try:
        text = recognizer.recognize_google(audio, language=language)
        return text
    except sr.UnknownValueError:
        return "无法理解语音"
    except sr.RequestError:
        return "服务出错"
def translate_text(text, src_lang, dest_lang):
    translator = Translator()
    translation = translator.translate(text, src=src_lang, dest=dest_lang)
    return translation.text
def main():
    recognizer = sr.Recognizer()
    microphone = sr.Microphone()
    # 假设客户使用西班牙语
    customer_speech = recognize_customer_speech(recognizer, microphone, language="es")
    print("客户说：", customer_speech)
    # 将客户的西班牙语翻译成英语
    translated_speech = translate_text(customer_speech, "es", "en")
    print("翻译成英语：", translated_speech)
    # 这里可以根据翻译后的内容回应客户
if __name__ == "__main__":
    main()