pip install SpeechRecognition gTTS PyAudio

conda install SpeechRecognition gTTS PyAudio

import pyaudio
 
# 初始化PyAudio
p = pyaudio.PyAudio()
 
# 打開(kāi)音頻流
stream = p.open(format=pyaudio.paInt16,
                channels=1,
                rate=44100,
                input=True,
                frames_per_buffer=1024)
 
print("開(kāi)始錄音...")
 
# 錄制音頻
frames = []
for _ in range(0, int(44100 / 1024 * 5)):  # 錄制5秒
    data = stream.read(1024)
    frames.append(data)
 
print("錄音結(jié)束")
 
# 關(guān)閉音頻流
stream.stop_stream()
stream.close()
p.terminate()
 
# 將音頻數(shù)據(jù)保存為WAV文件
with open('recording.wav', 'wb') as wf:
    wf.write(b''.join(frames))

import speech_recognition as sr
 
# 初始化識(shí)別器
r = sr.Recognizer()
 
# 加載音頻文件
with sr.AudioFile('recording.wav') as source:
    audio_data = r.record(source)  # 讀取整個(gè)音頻文件
 
# 使用Google語(yǔ)音識(shí)別API進(jìn)行識(shí)別
try:
    text = r.recognize_google(audio_data, language='zh-CN')  # 中文識(shí)別
    print("識(shí)別結(jié)果: " + text)
except sr.UnknownValueError:
    print("無(wú)法識(shí)別音頻")
except sr.RequestError as e:
    print("請(qǐng)求錯(cuò)誤; {0}".format(e))

import tkinter as tk
from tkinter import messagebox
import speech_recognition as sr
from gtts import gTTS
import os
import pyaudio
 
class VoiceMemoApp:
    def __init__(self, root):
        self.root = root
        self.root.title("智能語(yǔ)音備忘錄")
        
        # 創(chuàng)建界面元素
        self.record_button = tk.Button(root, text="開(kāi)始錄音", command=self.start_recording)
        self.record_button.pack()
        
        self.text_label = tk.Label(root, text="識(shí)別結(jié)果將顯示在這里")
        self.text_label.pack()
        
        self.save_button = tk.Button(root, text="保存?zhèn)渫?, command=self.save_memo)
        self.save_button.pack()
        
        self.p = pyaudio.PyAudio()
        self.stream = None
        self.frames = []
        
    def start_recording(self):
        self.record_button.config(text="錄音中...", state=tk.DISABLED)
        self.stream = self.p.open(format=pyaudio.paInt16,
                                  channels=1,
                                  rate=44100,
                                  input=True,
                                  frames_per_buffer=1024)
        self.root.after(5000, self.stop_recording)  # 5秒后自動(dòng)停止錄音
        
    def stop_recording(self):
        self.stream.stop_stream()
        self.stream.close()
        self.record_button.config(text="開(kāi)始錄音", state=tk.NORMAL)
        self.recognize_speech()
        
    def recognize_speech(self):
        r = sr.Recognizer()
        with sr.AudioFile('temp_recording.wav', 'wb') as f:
            f.write(b''.join(self.frames))
        with sr.AudioFile('temp_recording.wav') as source:
            audio_data = r.record(source)
        try:
            text = r.recognize_google(audio_data, language='zh-CN')
            self.text_label.config(text=text)
        except sr.UnknownValueError:
            messagebox.showerror("錯(cuò)誤", "無(wú)法識(shí)別音頻")
        except sr.RequestError as e:
            messagebox.showerror("錯(cuò)誤", f"請(qǐng)求錯(cuò)誤: {e}")
        
    def save_memo(self):
        text = self.text_label.cget("text")
        if text == "識(shí)別結(jié)果將顯示在這里":
            messagebox.showwarning("警告", "請(qǐng)先錄制并識(shí)別語(yǔ)音")
            return
        # 保存為文本文件
        with open("memo.txt", "a") as f:
            f.write(text + "\n")
        # 生成語(yǔ)音文件
        tts = gTTS(text, lang='zh-cn')
        tts.save("memo.mp3")
        messagebox.showinfo("成功", "備忘錄已保存為memo.txt和memo.mp3")
 
if __name__ == "__main__":
    root = tk.Tk()
    app = VoiceMemoApp(root)
    root.mainloop()