语音机器人

Posted 2020-12-12 hrunjie

  1 from aip import AipSpeech
  2 import requests
  3 import json
  4 import speech_recognition as sr
  5 import win32com.client
  6 from selenium import webdriver
  7 import time
  8 
  9 
 10  
 11 speaker = win32com.client.Dispatch("SAPI.SpVoice")
 12 
 13 def watch_movie(a):
 14     driver.get(‘https://www.iqiyi.com‘)
 15     time.sleep(2)
 16 
 17     movie = driver.find_element_by_id(‘J-header-search-input‘)
 18     movie.send_keys(a)
 19 
 20 
 21 def my_record(rate=16000):
 22     r = sr.Recognizer()
 23     with sr.Microphone(sample_rate=rate) as source:
 24         print("please say something")
 25         audio = r.listen(source)
 26  
 27     with open("myvoices.wav", "wb") as f:
 28         f.write(audio.get_wav_data())
 29  
 30  
 31 APP_ID = ‘20315758‘
 32 API_KEY = ‘pPMQbHo7ccNrGo7XvLjyupEi‘
 33 SECRET_KEY = ‘MfFOtlK8m9SFUIlrUpCRl7bZl65datG3‘
 34 client = AipSpeech(APP_ID, API_KEY, SECRET_KEY)
 35 path = ‘myvoices.wav‘
 36  
 37  
 38 def listen():
 39     with open(path, ‘rb‘) as fp:
 40         voices = fp.read()
 41     try:
 42         result = client.asr(voices, ‘wav‘, 16000, {‘dev_pid‘: 1537, })
 43         result_text = result["result"][0]
 44         print("我  : " + result_text)
 45         return result_text
 46     except KeyError:
 47         print("KeyError")
 48         speaker.Speak("我没有听清楚，请再说一遍...")
 49  
 50  
 51 turing_api_key = "f502483535bb4f9eac11b822d9307b06"
 52 api_url = "http://openapi.tuling123.com/openapi/api/v2" 
 53 headers = {‘Content-Type‘: ‘application/json;charset=UTF-8‘}
 54  
 55  
 56 def Turing(text_words=""):
 57     req = {
 58         "reqType": 0,
 59         "perception": {
 60             "inputText": {
 61                 "text": text_words
 62             },
 63  
 64             "selfInfo": {
 65                 "location": {
 66                     "city": "潮州",
 67                     "province": "广东",
 68                     "street": "潮安"
 69                 }
 70             }
 71         },
 72         "userInfo": {
 73             "apiKey": turing_api_key,  # 你的图灵机器人apiKey
 74             "userId": "Nieson"  # 用户唯一标识(随便填, 非密钥)
 75         }
 76     }
 77  
 78     req["perception"]["inputText"]["text"] = text_words
 79     response = requests.request("post", api_url, json=req, headers=headers)
 80     response_dict = json.loads(response.text)
 81  
 82     result = response_dict["results"][0]["values"]["text"]
 83     print("大白: " + result)
 84     print(‘........‘)
 85     return result
 86  
 87  
 88 print(‘.........‘)
 89 while True:
 90     my_record()
 91     request = listen()
 92     response = Turing(request)
 93     speaker.Speak(response)
 94     if request==‘再见。‘:
 95         break
 96     if request[:5]==‘打开爱奇艺‘:
 97         movie = request[6:-1]
 98         driver = webdriver.Chrome()
 99         driver.get(‘https://so.iqiyi.com/so/q_‘+movie+‘?source=default&sr=1348372637207‘)
100         time.sleep(3)
101 
102 
103 
104 
105 
106 print(‘  
欢迎下次使用!‘)
107  

 

1.基本原理：

（1）语音生成音频文件，录音并以当前时间保存到voices文件中

（2）音频文件转文字：采用百度语音识别

（3）与机器人对话：调用图灵机器人，回复对话

（4）用requests库，爬取机器人的回复，输出机器人回答（回复‘再见’时，结束程序）

 

2.用到的库 ：api （图灵机器人，百度语音），requests ，json （爬虫），

speech_recognition（语音识别）