Chatbot

Speech to text

https://speechnotes.co/

Text-to-Speech

收費項目範例

https://ai.baidu.com/tech/speech/tts

輕鬆導入微軟Azure語音辨識API

Amazon Polly逼真語音 | 將文字轉化為逼真語音‎

Google Speech API | 語音轉文字 | google.com‎

不收費項目

TTS Web sample

Baidu sample

python sample

>>> import webbrowser

>>> webbrowser.open('http://translate.google.com/translate_tts?ie=UTF-8&total=1&idx=0&textlen=32&client=tw-ob&q=%E8%83%B0%E5%B3%B6%E7%B4%A0%E9%96%8B%E5%A7%8B%20%E8%83%B0%E5%B3%B6%E7%B4%A0%E6%B3%A8%E5%B0%84%E7%B5%90%E6%9D%9F&tl=zh-TW')

gTTS (Google Text-to-Speech)

Installation

pip3 install gTTS

pip3 install gTTS-token --upgrade

Quickstart

Command Line:

$ gtts-cli 'hello' --output hello.mp3

Module:

>>> from gtts import gTTS

>>> tts = gTTS('hello')

>>> tts.save('hello.mp3')

mp3 play with pygame

pip install pygame

使用暫存檔儲存mp3

from time import sleep

import tempfile

from gtts import gTTS

from pygame import mixer

def speak(sentence):

with tempfile.NamedTemporaryFile(delete=True) as fp:

print(sentence)

tts = gTTS(text=sentence, lang='zh-TW')

tts.save('{}.mp3'.format(fp.name))

mixer.init()

mixer.music.load('{}.mp3'.format(fp.name))

mixer.music.play()

while mixer.music.get_busy():

sleep(1)

print ("done")

中文 sample

from time import sleep

import tempfile

from gtts import gTTS

from pygame import mixer

def speak(sentence):

with tempfile.NamedTemporaryFile(delete=True) as fp:

tts = gTTS(text=sentence, lang='zh-TW')

tts.save('{}.mp3'.format(fp.name))

mixer.init()

mixer.music.load('{}.mp3'.format(fp.name))

mixer.music.play(1)

while mixer.music.get_busy():

sleep(1)

print ("done")

speak('臺北醫學大學老山山教授')

from gtts import gTTS

from playsound import playsound

mytext="你好，我是老山山教授"

language='ZH-TW'

myobj=gTTS(text=mytext,lang=language)

myobj.save("c:/Users/jjpc3/Desktop/welcome1.mp3")

playsound("c:/Users/jjpc3/Desktop/welcome1.mp3")

範例：每分鐘報時系統

import time

from gtts import gTTS

import tempfile

from pygame import mixer

from datetime import datetime

from time import strftime

def say(text, filename=None):

with tempfile.NamedTemporaryFile(delete=True) as temp:

tts = gTTS(text, lang='zh-TW',slow=False)

if filename is None:

filename = "{}.mp3".format(temp.name)

tts.save(filename)

mixer.init()

mixer.music.load(filename)

mixer.music.play()

while mixer.music.get_busy() == True:

continue

mixer.quit()

while(1):

hour=datetime.now().strftime('%H')

minute=datetime.now().strftime('%M')

second=datetime.now().strftime('%S')

print(second)

while second == '00':

say("現在是"+hour+"點"+minute+"分")

second=datetime.now().strftime('%S')

time.sleep(1)

範例：提醒系統

import time

from gtts import gTTS

import tempfile

from pygame import mixer

from datetime import datetime

from time import strftime

remind = {

'1332':'主人該喝水了',

'1333':'主人你該吃藥了',

'1334':'謝謝你的讚美'

}

def say(text, filename=None):

with tempfile.NamedTemporaryFile(delete=True) as temp:

tts = gTTS(text, lang='zh-TW',slow=False)

if filename is None:

filename = "{}.mp3".format(temp.name)

tts.save(filename)

mixer.init()

mixer.music.load(filename)

mixer.music.play()

while mixer.music.get_busy() == True:

continue

mixer.quit()

while(1):

hour=datetime.now().strftime('%H')

minute=datetime.now().strftime('%M')

second=datetime.now().strftime('%S')

print(hour+minute+second)

while second == '00':

if remind.get(hour+minute,'none') != 'none':

print(remind.get(hour+minute))

say("現在是"+hour+"點"+minute+"分"+remind.get(hour+minute))

time.sleep(1)

second=datetime.now().strftime('%S')

time.sleep(1)

範例：定時控制系統

import time

from gtts import gTTS

import tempfile

from pygame import mixer

from datetime import datetime

from time import strftime

from urllib.request import urlopen

import urllib

import urllib.parse

import json

remind = {

'1951':'http://61.219.106.180/xxxxx/commexec_json.php?user=jj&lang=cmn-Hant-TW&command=%E6%89%93%E9%96%8B%E5%B0%8F%E5%A4%9C%E7%87%88',

'1952':'http://61.219.106.180/Dxxxx/commexec_json.php?user=jj&lang=cmn-Hant-TW&command='+urllib.parse.quote("關閉小夜燈"),

'1953':'http://13.56.153.188/Drxxxx/IOcontrol.php?device=2nwZTtSl&command=toggle'

}

def say(text, filename=None):

with tempfile.NamedTemporaryFile(delete=True) as temp:

tts = gTTS(text, lang='zh-TW',slow=False)

if filename is None:

filename = "{}.mp3".format(temp.name)

tts.save(filename)

mixer.init()

mixer.music.load(filename)

mixer.music.play()

while mixer.music.get_busy() == True:

continue

mixer.quit()

while(1):

hour=datetime.now().strftime('%H')

minute=datetime.now().strftime('%M')

second=datetime.now().strftime('%S')

print(hour+minute+second)

while second == '00':

if remind.get(hour+minute,'none') != 'none':

print(remind.get(hour+minute))

u = urlopen(remind.get(hour+minute))

say("現在是"+hour+"點"+minute+"分正在執行你的指令")

time.sleep(1)

second=datetime.now().strftime('%S')

time.sleep(1)

speech_recognition

https://pypi.org/project/SpeechRecognition/

pip3 install SpeechRecognition

PyAudio (for microphone users)

raspberry pi

sudo apt-get install python-pyaudio python3-pyaudio

sudo apt-get install flac

PC use

pip install pyaudio

or

python -m pip install pyaudio

or

pip install pipwin

pipwin install pyaudio

or add

brew install portaudio

brew link portaudio

範例：語音辨識

import speech_recognition as sr

#obtain audio from the microphone

def listenTo():

r=sr.Recognizer()

with sr.Microphone() as source:

print("Please wait. Calibrating microphone...")

#listen for 2 seconds and create the ambient noise energy level

r.adjust_for_ambient_noise(source, duration=2)

print("Say something!")

audio=r.listen(source)

# recognize speech using Google Speech Recognition

try:

print("JJ robot will answer your question:")

print(r.recognize_google(audio, language="zh-TW"))

return r.recognize_google(audio, language="zh-TW")

except sr.UnknownValueError:

print("Google Speech Recognition could not understand audio")

except sr.RequestError as e:

print("No response from Google Speech Recognition service: {0}".format(e))

listenTo()

Chatbot

範例：聊天機器人

import speech_recognition as sr

#obtain audio from the microphone

def listenTo():

r=sr.Recognizer()

with sr.Microphone() as source:

print("Please wait. Calibrating microphone...")

#listen for 2 seconds and create the ambient noise energy level

r.adjust_for_ambient_noise(source, duration=2)

print("Say something!")

audio=r.listen(source)

# recognize speech using Google Speech Recognition

try:

print("JJ robot will answer your question:")

# print(r.recognize_google(audio, language="zh-TW"))

return r.recognize_google(audio, language="zh-TW")

except sr.UnknownValueError:

print("Google Speech Recognition could not understand audio")

except sr.RequestError as e:

print("No response from Google Speech Recognition service: {0}".format(e))

from time import sleep

import tempfile

from gtts import gTTS

from pygame import mixer

def speak(sentence):

with tempfile.NamedTemporaryFile(delete=True) as fp:

tts = gTTS(text=sentence, lang='zh-TW')

tts.save('{}.mp3'.format(fp.name))

mixer.init()

mixer.music.load('{}.mp3'.format(fp.name))

mixer.music.play()

while mixer.music.get_busy():

sleep(1)

print ("done")

qa = {

'你好':'我很好',

'你是誰':'我是你的小管家',

'你很聰明':'謝謝你的讚美'

}

speak(qa.get(listenTo(), '我聽不懂'))

Translate

https://github.com/ssut/py-googletrans

pip3 install googletrans

pip3 install git+https://github.com/BoseCorp/py-googletrans.git --upgrade

範例：翻譯機器人

import speech_recognition as sr

from googletrans import Translator

translator = Translator()

#obtain audio from the microphone

def listenTo():

r=sr.Recognizer()

with sr.Microphone() as source:

print("Please wait. Calibrating microphone...")

#listen for 5 seconds and create the ambient noise energy level

r.adjust_for_ambient_noise(source, duration=1)

print("Say something!")

audio=r.listen(source)

# recognize speech using Google Speech Recognition

try:

print("JJ robot will answer your question:")

print(r.recognize_google(audio, language="zh-TW"))

#sentence = translator.translate(r.recognize_google(audio, language="zh-TW"), dest='en').text

return translator.translate(r.recognize_google(audio, language="zh-TW"), dest='ja').text

except sr.UnknownValueError:

print("Google Speech Recognition could not understand audio")

return " "

except sr.RequestError as e:

print("No response from Google Speech Recognition service: {0}".format(e))

return " "

from time import sleep

import tempfile

from gtts import gTTS

from pygame import mixer

def speak(sentence):

if sentence != " ":

with tempfile.NamedTemporaryFile(delete=True) as fp:

print(sentence)

tts = gTTS(text=sentence, lang='ja')

tts.save('{}.mp3'.format(fp.name))

mixer.init()

mixer.music.load('{}.mp3'.format(fp.name))

mixer.music.play()

while mixer.music.get_busy():

sleep(1)

print ("done")

import speech_recognition as sr

from googletrans import Translator

translator = Translator()

#obtain audio from the microphone

def listenTo():

r=sr.Recognizer()

with sr.Microphone() as source:

print("Please wait. Calibrating microphone...")

#listen for 5 seconds and create the ambient noise energy level

r.adjust_for_ambient_noise(source, duration=1)

print("Say something!")

audio=r.listen(source)

# recognize speech using Google Speech Recognition

try:

print("JJ robot will answer your question:")

print(r.recognize_google(audio, language="zh-TW"))

#sentence = translator.translate(r.recognize_google(audio, language="zh-TW"), dest='en').text

return translator.translate(r.recognize_google(audio, language="zh-TW"), dest='ja').text

except sr.UnknownValueError:

print("Google Speech Recognition could not understand audio")

return " "

except sr.RequestError as e:

print("No response from Google Speech Recognition service: {0}".format(e))

return " "

from time import sleep

import tempfile

from gtts import gTTS

from pygame import mixer

def speak(sentence):

if sentence != " ":

with tempfile.NamedTemporaryFile(delete=True) as fp:

print(sentence)

tts = gTTS(text=sentence, lang='ja')

tts.save('{}.mp3'.format(fp.name))

mixer.init()

mixer.music.load('{}.mp3'.format(fp.name))

while True:

speak(listenTo())

Google 查詢機器人

Beautiful Soup 本身只是一個 HTML 解析工具，它並不負責下載網頁，所以通常我們在開發爬蟲程式時，會搭配 requests 模組一同使用。

import requests

from bs4 import BeautifulSoup

範例：Google 查詢機器人

import speech_recognition as sr

#obtain audio from the microphone

def listenTo():

r=sr.Recognizer()

with sr.Microphone() as source:

print("Please wait. Calibrating microphone...")

#listen for 5 seconds and create the ambient noise energy level

r.adjust_for_ambient_noise(source, duration=1)

print("Say something!")

audio=r.listen(source)

# recognize speech using Google Speech Recognition

try:

print("JJ robot will answer your question:")

answertxt = r.recognize_google(audio, language="zh-TW")

print(answertxt)

return answertxt

except sr.UnknownValueError:

print("Google Speech Recognition could not understand audio")

return " "

except sr.RequestError as e:

print("No response from Google Speech Recognition service: {0}".format(e))

return " "

from time import sleep

import tempfile

from gtts import gTTS

from pygame import mixer

def speak(sentence):

with tempfile.NamedTemporaryFile(delete=True) as fp:

print(sentence)

if sentence == " ":

return

tts = gTTS(text=sentence, lang='zh-TW')

tts.save('{}.mp3'.format(fp.name))

mixer.init()

mixer.music.load('{}.mp3'.format(fp.name))

mixer.music.play()

while mixer.music.get_busy():

sleep(1)

print ("done")

import requests

from bs4 import BeautifulSoup

import urllib.parse

def ask_google(sentence):

if sentence == " ":

return " "

req1=requests.get('https://www.google.com.tw/search?source=lnt&tbas=0&sa=X&ved=0ahUKEwjo5PfQzc3cAhUP5rwKHU6CCvUQpwUIHg&biw=1920&bih=898&q='+urllib.parse.quote(sentence))

#print(req1.text)

soup = BeautifulSoup(req1.text,'html.parser')

divs = soup.find('div', 'g')

print(divs.text)

if len(divs.text) > 100:

req1=requests.get('https://www.google.com.tw/search?source=lnt&tbs=li:1&sa=X&ved=0ahUKEwjo5PfQzc3cAhUP5rwKHU6CCvUQpwUIHg&biw=1920&bih=898&q='+urllib.parse.quote(sentence))

soup = BeautifulSoup(req1.text,'html.parser')

divs = soup.find('div', 'g')

print(divs.text)

if len(divs.text) > 100:

return " "

return divs.text

return divs.text

while True:

speak(ask_google(listenTo()))

Baidu TTS

import os

import sys, locale

import tempfile

from time import sleep

from pygame import mixer # Load the required library

class Speak:

def __init__ (self,message,times,launguage,isRemove):

'''

传入参数 消息 次数 是否清除合成的语音文件

'''

self.tts(message,launguage)

#说几次

for i in range(0,times):

self.say()

if isRemove:

self.over()

def tts(self,message,launguage):

try:

import requests

except:

exit(-1)

import urllib

s = requests.Session()

mes=''

if launguage is "zh":

mes="http://tts.baidu.com/text2audio?spd=2&lan=zh&pid=101&vol=9&ie=UTF-8&text="

else:

mes="http://tts.baidu.com/text2audio?spd=2&lan=en&pid=101&vol=9&ie=UTF-8&text="

s.get(mes+ urllib.request.quote(message))

res = s.get(mes+ urllib.request.quote(message)).content

f = open("tts-temp.mp3", "wb+")

f.write(res)

f.close()

def say(self):

mixer.init()

mixer.music.load('tts-temp.mp3')

mixer.music.play()

while mixer.music.get_busy():

sleep(0.05)

mixer.music.stop()

mixer.quit()

def over(self):

os.system("del tts-temp.mp3")

def play(music):

os.system("play "+music)

man=Speak("我是臺北醫學大學學生",1,"zh",True)

把 "我是臺北醫學大學學生" 改成 sys.argv[1] 就可以成為一個 API

Baidu 每分鐘報時系統

import os

import time

from datetime import datetime

from time import strftime

def speak(sentence):

print(sentence)

os.system("baidu_TTS '"+sentence+"'" )

while(1):

hour=datetime.now().strftime('%H')

minute=datetime.now().strftime('%M')

second=datetime.now().strftime('%S')

print(second)

while second == '00':

speak("現在是"+hour+"點"+minute+"分")

second=datetime.now().strftime('%S')

time.sleep(1)

Baidu Chatbot

import os

import speech_recognition as sr

#obtain audio from the microphone

def listenTo():

r=sr.Recognizer()

with sr.Microphone() as source:

print("Please wait. Calibrating microphone...")

#listen for 5 seconds and create the ambient noise energy level

r.adjust_for_ambient_noise(source, duration=5)

print("Say something!")

audio=r.listen(source)

# recognize speech using Google Speech Recognition

try:

print("JJ robot will answer your question:")

# print(r.recognize_google(audio, language="zh-TW"))

return r.recognize_google(audio, language="zh-TW")

except sr.UnknownValueError:

print("Google Speech Recognition could not understand audio")

except sr.RequestError as e:

print("No response from Google Speech Recognition service: {0}".format(e))

from time import sleep

def speak(sentence):

print(sentence)

os.system("baidu_TTS '"+sentence+"'" )

qa = {

'你好':'我很好',

'你是誰':'我是你的小管家',

'你很聰明':'謝謝你的讚美'

}

speak(qa.get(listenTo(), '我聽不懂'))

SpeechCognition_sendMQTT

必須安裝 pip3 install paho-mqtt

import speech_recognition as sr

import paho.mqtt.client as mqtt

import paho.mqtt.publish as publish

import time

HOST = "120.97.62.97"

PORT = 1883

def on_connect(client, userdata, flags, rc):

print("Connected with result code "+str(rc))

client.subscribe("AI/speech_reply/user")

def on_message(client, userdata, msg):

print(msg.topic+" "+msg.payload.decode("utf-8"))

def send_message(user, msg, device):

# print(msg)

client_id = time.strftime('%Y%m%d%H%M%S',time.localtime(time.time()))

client = mqtt.Client(client_id) # ClientId不能重复，所以使用当前时间

client.username_pw_set("drchiu", "8888") # 必须设置，否则会返回「Connected with result code 4」

client.on_connect = on_connect

client.on_message = on_message

client.connect(HOST, PORT, 60)

client.publish("AI/speech_reply/user", '{"user":"'+user+'","text":"'+msg+'","device":"'+device+'","lang":"cmn-Hant-TW"}', qos=0, retain=False)

if __name__ == '__main__':

while True:

#obtain audio from the microphone

r=sr.Recognizer()

with sr.Microphone() as source:

print("Please wait. Calibrating microphone...")

#listen for 5 seconds and create the ambient noise energy level

r.adjust_for_ambient_noise(source, duration=1)

print("Say something!")

audio=r.listen(source)

# recognize speech using Google Speech Recognition

try:

print("Google Speech Recognition thinks you said:")

print(r.recognize_google(audio, language="zh-TW"))

send_message('jj',r.recognize_google(audio, language="zh-TW"),'all')

except sr.UnknownValueError:

print("Google Speech Recognition could not understand audio")

except sr.RequestError as e:

print("No response from Google Speech Recognition service: {0}".format(e))

MQTT_speech

from gtts import gTTS

import tempfile

from pygame import mixer

import paho.mqtt.client as mqtt

import paho.mqtt.publish as publish

import paho.mqtt.subscribe as subscribe

import time

import json

HOST = "120.97.62.97"

PORT = 1883

ID = "raspberry1"

def say(text, filename=None):

with tempfile.NamedTemporaryFile(delete=True) as temp:

tts = gTTS(text, lang='zh-TW',slow=False)

if filename is None:

filename = "{}.mp3".format(temp.name)

tts.save(filename)

mixer.init()

mixer.music.load(filename)

mixer.music.play()

while mixer.music.get_busy() == True:

continue

mixer.quit()

def on_connect(client, userdata, flags, rc):

print("Connected with result code "+str(rc))

client.subscribe("AI/speech_reply/user")

def on_message(client, userdata, msg):

x=msg.payload.decode("utf-8")

print(x)

y = json.loads(x)

print(y["text"])

print(msg.topic+" "+y["text"])

say(y["text"])

def send_message(user, msg, device):

# print(msg)

client_id = time.strftime('%Y%m%d%H%M%S',time.localtime(time.time()))

client = mqtt.Client(client_id) # ClientId不能重复，所以使用当前时间

client.username_pw_set("drchiu", "8888") # 必须设置，否则会返回「Connected with result code 4」

client.on_connect = on_connect

client.on_message = on_message

client.connect(HOST, PORT, 60)

client.publish("AI/speech_reply/user", '{"user":"'+user+'","text":"'+msg+'","device":"'+device+'","lang":"cmn-Hant-TW"}', qos=0, retain=False)

if __name__ == '__main__':

client_id = time.strftime('%Y%m%d%H%M%S',time.localtime(time.time()))

client = mqtt.Client(client_id) # ClientId不能重复，所以使用当前时间

client.username_pw_set("drchiu", "8888")

######Bind function to callback

client.on_message=on_message

#####

print("connecting to broker ",HOST )

client.connect(HOST )#connect

client.loop_start() #start loop to process received messages

print("subscribing ")

client.subscribe("all/speech_reply")#subscribe

#client.subscribe(ID+"/speech_reply", qos=1)

#client.message_callback_add("all/speech_reply", on_message )

#client.loop_forever()

#print("publishing ")

#client.publish("all/speech_reply","on")#publish

#client.disconnect() #disconnect

#client.loop_stop() #stop loop

say("主人你好，很高興能為你服務")

while True:

subscribe.callback(on_message, ID+"/speech_reply", hostname=HOST )

#subscribe.message_callback_add(on_message, "all/speech_reply", hostname=HOST )

jj_speech_recognition.py

GenJyuuGothicX-Normal.ttf doll_gif.zip

import requests

import json

import speech_recognition as sr

from time import sleep

import tempfile

from gtts import gTTS

from pygame import mixer

import subprocess

import pygame

import datetime

import paho.mqtt.client as mqtt

import paho.mqtt.publish as publish

import paho.mqtt.subscribe as subscribe

import time

##from concurrent.futures import ThreadPoolExecutor

import threading

HOST = "120.97.62.97"

PORT = 1883

ID = "raspberry1"

url="http://61.219.106.180/drchiu/commexec_doll.php"

user = 'jj'

#obtain audio from the microphone

def listenTo():

r=sr.Recognizer()

with sr.Microphone() as source:

print("Please wait. Calibrating microphone...")

#listen for 1 seconds and create the ambient noise energy level

r.adjust_for_ambient_noise(source, duration=1)

print("我正在聆聽你的命令!")

show_ear()

#subprocess.Popen("sudo python3 ws2812.py", shell=True)

audio=r.listen(source)

# recognize speech using Google Speech Recognition

try:

print("JJ robot listen sentence from you:")

answer = r.recognize_google(audio, language="zh-TW")

print(answer)

return answer

except sr.UnknownValueError:

print("Google Speech Recognition could not understand audio")

return "again"

except sr.RequestError as e:

print("No response from Google Speech Recognition service: {0}".format(e))

def speak(sentence):

with tempfile.NamedTemporaryFile(delete=True) as fp:

tts = gTTS(text=sentence, lang='zh-TW')

tts.save('{}.mp3'.format(fp.name))

mixer.init()

mixer.music.load('{}.mp3'.format(fp.name))

mixer.music.play()

while mixer.music.get_busy() == True:

continue

#print ("done")

mixer.quit()

def ask_cloud(text):

r = requests.get(url, params={'user': user,'command':text,'reply':'json'})

data = r.json()

print(data)

example = data["reply"]

print(example)

screen_show(text,example)

return example

def screen_show(text1, text2):

font = pygame.font.Font("GenJyuuGothicX-Normal.ttf",36)

text = font.render( text2, True, (0, 128, 0))

font = pygame.font.Font("GenJyuuGothicX-Normal.ttf", 76)

text1 = font.render( text1, True, (0, 0, 128))

for event in pygame.event.get():

if event.type == pygame.QUIT:

done = True

if event.type == pygame.KEYDOWN and event.key == pygame.K_ESCAPE:

done = True

screen.fill((255, 255, 255))

screen.blit(text1,

(400 - text1.get_width() // 2, 240- text1.get_height() // 2))

screen.blit(text,

(400 - text.get_width() // 2, 240 - text.get_height() // 2 + 120))

image = pygame.image.load("eyes.png")

image.convert()

screen.blit(image, (250,70))

pygame.display.flip()

pygame.display.flip()

def on_connect(client, userdata, flags, rc):

print("Connected with result code "+str(rc))

client.subscribe("AI/speech_reply/user")

def on_message(client, userdata, msg):

x=msg.payload.decode("utf-8")

print(x)

y = json.loads(x)

print(y["text"])

##print(msg.topic+" "+y["text"])

speak(y["text"])

def send_message(user, msg, device):

# print(msg)

client_id = time.strftime('%Y%m%d%H%M%S',time.localtime(time.time()))

client = mqtt.Client(client_id) # ClientId不能重复，所以使用当前时间

client.username_pw_set("drchiu", "8888") # 必须设置，否则会返回「Connected with result code 4」

client.on_connect = on_connect

client.on_message = on_message

client.connect(HOST, PORT, 60)

client.publish("AI/speech_reply/user", '{"user":"'+user+'","text":"'+msg+'","device":"'+device+'","lang":"cmn-Hant-TW"}', qos=0, retain=False)

##def process():

##with ThreadPoolExecutor() as executor:

##robot()

##mqtt_net()

def mqtt_net():

client_id = time.strftime('%Y%m%d%H%M%S',time.localtime(time.time()))

client = mqtt.Client(client_id) # ClientId不能重复，所以使用当前时间

client.username_pw_set("drchiu", "8888")

###Bind function to callback

client.on_message=on_message

#####

print("connecting to broker ",HOST )

client.connect(HOST )#connect

client.loop_start() #start loop to process received messages

print("subscribing ")

client.subscribe("all/speech_reply")#subscribe

#client.subscribe(ID+"/speech_reply", qos=1)

#client.message_callback_add("all/speech_reply", on_message )

#client.loop_forever()

while True:

subscribe.callback(on_message, ID+"/speech_reply", hostname=HOST )

def show_ear():

image = pygame.image.load("ear.png")

image.convert()

screen.blit(image, (680,20))

image = pygame.image.load("ear_left.png")

image.convert()

screen.blit(image, (20,20))

image = pygame.image.load("eyes.png")

image.convert()

screen.blit(image, (250,70))

pygame.display.flip()

def robot():

wait_count=1

##screen = pygame.display.set_mode((800, 480))

##screen = pygame.display.set_mode((0, 0), pygame.FULLSCREEN)

screen_show('','主人你好,很高興能為你服務 ')

speak('主人你好,很高興能為你服務 ')

#subprocess.Popen("sudo python3 MQTT_speech.py", shell=True)

##subscribe.callback(on_message, ID+"/speech_reply", hostname=HOST )

while True:

#show_ear()

answer = listenTo()

if (answer == 'again'):

screen_show('', '請下達你的指令')

wait_count =wait_count +1

if(wait_count >30):

speak('主人你好')

wait_count =1

else:

text1=ask_cloud(answer)

if(text1 != '我不懂你的意思'):

speak(text1)

###

if __name__ == '__main__':

pygame.init()

#screen = pygame.display.set_mode((0, 0), pygame.FULLSCREEN)

screen = pygame.display.set_mode((640, 480))

t = threading.Thread(target = mqtt_net)

# 執行該子執行緒

t.start()

robot()

pocketsphinx

sudo apt-get install pulseaudio

Google Cloud text to speech

set GOOGLE_APPLICATION_CREDENTIALS=jjlab-speech-send-cc165e9881e4.json

"""Google Cloud Text-To-Speech API sample application .

Example usage:

python text2speech.py 這是一個文字轉語音範例

"""

from pygame import mixer

import tempfile

from time import sleep

import sys

def text2speech(txt):

from google.cloud import texttospeech

# Instantiates a client

client = texttospeech.TextToSpeechClient()

# Set the text input to be synthesized

synthesis_input = texttospeech.types.SynthesisInput(text= txt)

# Build the voice request, select the language code ("en-US") and the ssml

# voice gender ("neutral")

voice = texttospeech.types.VoiceSelectionParams(

language_code='zh-hant',

ssml_gender=texttospeech.enums.SsmlVoiceGender.NEUTRAL)

# Select the type of audio file you want returned

audio_config = texttospeech.types.AudioConfig(

audio_encoding=texttospeech.enums.AudioEncoding.MP3)

# Perform the text-to-speech request on the text input with the selected

# voice parameters and audio file type

response = client.synthesize_speech(synthesis_input, voice, audio_config)

# [END tts_quickstart]

with tempfile.NamedTemporaryFile(delete=True) as fp:

# The response's audio_content is binary.

with open('{}.mp3'.format(fp.name), 'wb') as out:

# Write the response to the output file.

out.write(response.audio_content)

print('Audio content written to file "output.mp3"')

#print(text)

#tts = gTTS(text=sentence, lang='zh-TW')

#tts.save('{}.mp3'.format(fp.name))

mixer.init()

mixer.music.load('{}.mp3'.format(fp.name))

mixer.music.play()

while mixer.music.get_busy():

sleep(1)

print("done")

if __name__ == '__main__':

text2speech(sys.argv[1])

python text2speech.py 這是一個文字轉語音範例

Google cloud text2speech + MQTT

安裝 Google 套件

pip3 install gcloud

pip3 install google-api-python-client

pip3 install google-cloud-speech

pip3 install "google-cloud-texttospeech<2.0.0"

如果有黃色警告時加入

echo 'export PATH=/home/pi/.local/bin:$PATH' >>~/.bashrc

source ~/.bashrc

pip3 uninstall gcloud

pip3 install gcloud 重新安裝一次就好了

"""Google Cloud Text-To-Speech API sample application

Example usage:

windows: set GOOGLE_APPLICATION_CREDENTIALS=jjlab-speech-send-cc165e9881e4.json or

linux: export GOOGLE_APPLICATION_CREDENTIALS="jjlab-speech-send-cc165e9881e4.json"

python mqtt_google_say.py

MQTT:

topic jj/speech_reply

data {"text":"講個笑話給我聽","device":"windows1","lang":"cmn-Hant-TW"}

http://3.114.197.59/MQTT/publish_97.php?topic=jj/speech_reply&value={"text":"簡文山教授是我的偶像","device":"windows1","lang":"cmn-Hant-TW"}

"""

import pygame

from pygame import mixer

import tempfile

from time import sleep

import urllib.request

import sys

import paho.mqtt.client as mqtt

import paho.mqtt.publish as publish

import paho.mqtt.subscribe as subscribe

import time

import json

HOST = "120.97.62.97"

PORT = 1883

ID = "windows1"

group = "tmu_device"

lang = "cmn-Hant-TW"

user = "jj"

def play_song(url):

print('Beginning file download with urllib2...')

#url = 'http://hiot.prof-jj.com/music/clapping.mp3'

urllib.request.urlretrieve(url, './play.mp3')

pygame.init()

pygame.mixer.init()

screen=pygame.display.set_mode([1,1])

pygame.time.delay(1000)

MUSIC_END = pygame.USEREVENT+1

pygame.mixer.music.set_endevent(MUSIC_END)

pygame.mixer.music.load('./play.mp3')

pygame.mixer.music.play()

print("play mp3")

running = True

while running:

for event in pygame.event.get():

if event.type == MUSIC_END:

running = False

if running == False:

pygame.quit()

def say(txt,lang):

from google.cloud import texttospeech

# Instantiates a client

client = texttospeech.TextToSpeechClient()

# Set the text input to be synthesized

synthesis_input = texttospeech.types.SynthesisInput(text= txt)

# Build the voice request, select the language code ("en-US") and the ssml

# voice gender ("neutral") "MALE","FEMALE","NEUTRAL"

voice = texttospeech.types.VoiceSelectionParams(

language_code= lang,

ssml_gender=texttospeech.enums.SsmlVoiceGender.FEMALE)

# Select the type of audio file you want returned

audio_config = texttospeech.types.AudioConfig(

audio_encoding=texttospeech.enums.AudioEncoding.MP3)

# Perform the text-to-speech request on the text input with the selected

# voice parameters and audio file type

response = client.synthesize_speech(synthesis_input, voice, audio_config)

# [END tts_quickstart]

with tempfile.NamedTemporaryFile(delete=True) as fp:

# The response's audio_content is binary.

with open('{}.mp3'.format(fp.name), 'wb') as out:

# Write the response to the output file.

out.write(response.audio_content)

print('Audio content written to file "output.mp3"')

#print(text)

#tts = gTTS(text=sentence, lang='zh-TW')

#tts.save('{}.mp3'.format(fp.name))

mixer.init()

mixer.music.load('{}.mp3'.format(fp.name))

mixer.music.play()

while mixer.music.get_busy():

sleep(1)

print("done")

def on_connect(client, userdata, flags, rc):

print("Connected with result code "+str(rc))

client.subscribe("AI/speech_reply/user")

def on_message(client, userdata, msg):

x=msg.payload.decode("utf-8")

print(x)

y = json.loads(x)

if y["device"] == ID or y["device"] == group or y["device"] == "all_device":

if(y["text"][:1] != "/"):

print(y["text"])

print(msg.topic + " " + y["text"])

say(y["text"],y["lang"])

if(y["text"] == "/play_mp3"):

if(y["url"] != ""):

print(y["url"])

play_song(y["url"])

if(y["text"] == "/say_time"):

print(time.strftime("現在時間%H:%M",time.localtime()))

say(time.strftime("現在時間%H:%M",time.localtime()),y["lang"])

def send_message(user, msg, device):

# print(msg)

client_id = time.strftime('%Y%m%d%H%M%S',time.localtime(time.time()))

client = mqtt.Client(client_id) # ClientId不能重复，所以使用当前时间

client.username_pw_set("drchiu", "8888") # 必须设置，否则会返回「Connected with result code 4」

client.on_connect = on_connect

client.on_message = on_message

client.connect(HOST, PORT, 60)

client.publish("AI/speech_reply/user", '{"user":"'+user+'","text":"'+msg+'","device":"'+device+'","lang":"cmn-Hant-TW"}', qos=0, retain=False)

if __name__ == '__main__':

client_id = time.strftime('%Y%m%d%H%M%S',time.localtime(time.time()))

client = mqtt.Client(client_id) # ClientId不能重复，所以使用当前时间

client.username_pw_set("drchiu", "8888")

######Bind function to callback

client.on_message=on_message

#####

print("connecting to broker ",HOST )

client.connect(HOST )#connect

client.loop_start() #start loop to process received messages

print("subscribing ")

client.subscribe("all/speech_reply")#subscribe

#client.subscribe(ID+"/speech_reply", qos=1)

#client.message_callback_add("all/speech_reply", on_message )

#client.loop_forever()

#print("publishing ")

#client.publish("all/speech_reply","on")#publish

#client.disconnect() #disconnect

#client.loop_stop() #stop loop

say("主人你好，很高興能為你服務",lang)

while True:

subscribe.callback(on_message, user+"/speech_reply", hostname=HOST )

#subscribe.message_callback_add(on_message, "all/speech_reply", hostname=HOST )