def common_words_method(language_full, language):
welcome = (
"Welcome to the common words section. We are going to learn the %s translation of ten most common english words"
) % (language_full)
print(welcome)
f = open('data.txt', 'r')
random_indices = []
for i in range(5):
random_indices.append(random.randint(1, 3001))
#print(f[0],f[1])
line_number = 1
eng = []
esp = []
for line in f:
if (line_number in random_indices):
eng.append(line)
tran = translate_text_to_foreign(line, language)
esp.append(tran)
print(line + "->" + tran)
line_object_english = TextToSpeech(line, "en")
line_object = TextToSpeech(tran, language)
line_object_english.speech()
line_object.speech()
print(" ")
line_number += 1
print("Finally the words with their translations are- ")
for i in range(len(eng)):
print(eng[i], esp[i])
def __init__(self, detect_model="data/andrew2.net",
lyrics_model="data/keras_model_1200.h5",
lyrics_chars="data/chars.pkl"):
# microphone
self.mic = ResumableMicrophoneStream(SAMPLE_RATE, CHUNK_SIZE)
# wake word detector
self.detector = TriggerDetector(detect_model)
# speech and language services
self.speech_client = SpeechToText()
self.luis = LangUnderstand()
self.tts = TextToSpeech()
# lyrics generator model
self.lyrics_gen = LyricsGenerator(lyrics_model, lyrics_chars)
self.pred_queue = DetectQueue(maxlen=5)
self.is_wakeup = False
# pytft display
self.tft = TFTDisplay()
self.tft_queue = queue.Queue()
self.tft_thread = threading.Thread(target=self.tft_manage, args=())
self.tft_thread.daemon = True
self.tft_thread.start()
self.notify("hi_there")
def test_makeRequest(self):
face = ['will']
object = ['snowboard']
data = OneFaceOneObjectFake()
results = [
"",
"I am 100 percent sure that I can see a snowboard behind me on my left.",
"Yes, I am 100 percent sure that I can see exactly one snowboard behind me on my left.",
"I am 84 percent sure that I can see will. Not the real one though. behind me on my left.",
"Yes, I am 84 percent sure that I can see will. Not the real one though. behind me on my left.",
"I am 100 percent confident that the snowboard belongs to Will. behind me on my left.",
"",
"I am 100 percent sure that the snowboard is here behind me on my left.",
"Yes, I am 84 percent sure that a fake will is here, behind me on my left."
]
for id in range(1, 9):
if id == 6:
continue
data.instruction_id = id
text_to_speech = TextToSpeech(data,
'en-EN',
input_faces=face,
input_objects=object)
out = text_to_speech.callback()
self.assertEqual(out, results[id])
def start_rap(self):
media = vlc.MediaPlayer("assets/beat.mp3")
media.play()
media.set_time(20000)
for i in self.player_raps[self.current_player]:
TextToSpeech(i)
media.stop()
def test_makeRequest(self):
face = ['will']
object = ['snowboard']
data = ThreeFacesThreeObjectsTwoSameOneDifferent()
results = [
"", "I can see 2 snowboards, and 1 sports ball.",
"Yes, I can actually see more than one snowboard. I think there are 2. ",
"I can see will nico and eivinas.",
"Yes, I am 84 percent sure that I can see will behind me on my left.",
"I think that the following people have a snowboard: Will and Eivinas.",
"",
"I can actually see more than one snowboard. I think there are 2. One of them is here. ",
"I am 84 percent sure that will is here behind me on my left."
]
for id in range(1, 9):
if id == 6:
continue
data.instruction_id = id
text_to_speech = TextToSpeech(data,
'en-EN',
input_faces=face,
input_objects=object)
out = text_to_speech.callback()
self.assertEqual(out, results[id])
def test_makeRequest(self):
face = ['will']
object = ['snowboard']
data = ThreeFacesThreeObjectsThreeDifferent()
results = [
"", "I can see 1 sports ball, 1 snowboard, and 1 tennis racket.",
"Yes, I am 100 percent sure that I can see exactly one snowboard behind me on my left.",
"I can see will nico and eivinas.",
"Yes, I am 84 percent sure that I can see will behind me on my left.",
"I am 100 percent confident that the snowboard belongs to Will. behind me on my left.",
"",
"I am 100 percent sure that the snowboard is here behind me on my left.",
"I am 84 percent sure that will is here behind me on my left."
]
for id in range(1, 9):
if id == 6:
continue
data.instruction_id = id
text_to_speech = TextToSpeech(data,
'en-EN',
input_faces=face,
input_objects=object)
out = text_to_speech.callback()
self.assertEqual(out, results[id])
def __init__(self):
self.car = Vehicle()
# Verify that the Vehicle object was created
# print("Vehicle made!")
# self.find_own_pos("1362060585", "1362060062")
self.minute_warning = False
self.second_warning = False
self.meter_message = False
self.map = MapView()
self.text_to_speech = TextToSpeech()
def greeting(self):
try:
chat_bot = ChatBot(None)
response = chat_bot.send_message(self.first_message)
output = response[0]
self.con_id = response[1]
self.browser.append("<p>Agent: " + output + "</p>")
self.first_message = False
TextToSpeech(output)
except Exception as e:
print(e)
def send_text(message):
if message.text.lower() == 'привет':
bot.send_message(message.chat.id, 'Привет')
elif message.text.lower() == 'пока':
bot.send_message(message.chat.id, 'Прощай, создатель')
elif message.text.lower() == 'я тебя люблю':
bot.send_sticker(message.chat.id, 'CAADAgADZgkAAnlc4gmfCor5YbYYRAI')
else:
tts = TextToSpeech(text=message.text, lang='ru')
ogg_file = tts.save()
with open(ogg_file, 'rb') as f:
bot.send_voice(message.chat.id, f)
def receive_message(self, message):
try:
chat_bot = ChatBot(message)
if self.con_id is None:
response = chat_bot.send_message(self.first_message)
else:
response = chat_bot.send_message(self.first_message,
self.con_id)
output = response[0]
self.con_id = response[1]
self.browser.append("<p>Agent: " + output + "</p>")
self.first_message = False
TextToSpeech(output)
except Exception as e:
print(e)
def __init__(self, config):
# X. Initialize announcer.
self.announcer = TextToSpeech()
self.announcer.add_speech_text(
"Initializing system. Please wait for a moment.")
# X. Tf
self.broadcaster = tf.TransformBroadcaster()
self.listener = tf.TransformListener()
# Prepare shared object.
self.data = SharedData()
# Publishers & Subscribers
self.pubs = Publishers()
self.subs = Subscribers(self.announcer)
# Store tf
if (not config.pose_init):
while (self.subs.odom.get_object() is None):
time.sleep(0.5)
odom = self.subs.odom.get_object()
odom_xyzrpy = create_xyzrpy_from_pose(odom.pose.pose)
pose_stamped = PoseStamped()
pose_stamped.header.stamp = rospy.Time.now()
pose_stamped.header.frame_id = 'map'
pose_stamped.pose = create_pose_from_xyzrpy(config.xyzrpy)
self.data.last_valid_ndt_pose = pose_stamped
self.data.last_valid_tf_odom_to_map = create_transform_from_pose(
config.xyzrpy, odom_xyzrpy)
# Initilize Pose
self.initialize_pose(config)
# Start checking thread
self.checking_thread = threading.Thread(target=self.run_checking)
self.checking_thread.start()
# Start control
self.run_control()
# Terminate thread.
self.announcer.terminate()
self.checking_thread.join()
def execute(self, userdata):
"""
Execute function called in the state machine
Key arguments:
userdata -- state machine userdata object being passed around
"""
rospy.logdebug("Information")
rospy.loginfo(self.output_msg + "\n")
rospy.loginfo(userdata.recog_elements_in)
# Prepare sentence and get information on the camera
rospy.set_param(ACTIVE_CAMERA,
1) # Default front camera (changed by audio request)
text_to_speech = TextToSpeech(userdata.recog_elements_in,
userdata.lang_code_in,
userdata.action_in[1],
userdata.action_in[2])
self.output_msg = text_to_speech.callback()
# Indicating to the Display Controller that the display should show what the robot thinks
rospy.set_param(RECOG_UPPER_X, userdata.recog_elements_in.upper_x)
rospy.set_param(RECOG_UPPER_Y, userdata.recog_elements_in.upper_y)
rospy.set_param(RECOG_LOWER_X, userdata.recog_elements_in.lower_x)
rospy.set_param(RECOG_LOWER_Y, userdata.recog_elements_in.lower_y)
rospy.set_param(RECOG_LABEL, userdata.recog_elements_in.label)
rospy.set_param(RECOG_CAMERA, userdata.recog_elements_in.camera)
rospy.set_param(DISPLAY_INFORMATION, True)
# Maintain the thread for the duration of the speech out
audio_out_client.make_request(self.output_msg, userdata.lang_code_in)
# Resetting Parameters
rospy.set_param(RECOG_UPPER_X, [])
rospy.set_param(RECOG_UPPER_Y, [])
rospy.set_param(RECOG_LOWER_X, [])
rospy.set_param(RECOG_LOWER_Y, [])
rospy.set_param(RECOG_LABEL, [])
rospy.set_param(RECOG_CAMERA, [])
rospy.set_param(DISPLAY_INFORMATION, False)
return 'turning'
def test_makeRequest(self):
face = ['will']
object = ['snowboard']
data = EmptyData()
results = [
"", "I can’t see any objects.", "No, I cannot see any snowboard.",
"I can’t see anybody.", "No, I cannot see will.",
"I don’t think anybody has a snowboard.", "",
"I cannot see any snowboard.", "I cannot see will."
]
for id in range(1, 9):
if id == 6:
continue
data.instruction_id = id
text_to_speech = TextToSpeech(data,
'en-EN',
input_faces=face,
input_objects=object)
out = text_to_speech.callback()
self.assertEqual(out, results[id])
def __init__(self, config):
"""
Initialize
:param config: configuration
:type config: Config
"""
self.command_processor = CommandProcessor(
self._command_handlers(config.command_handlers))
self.robot = Robot(config.apiai.client_access_token,
config.apiai.language,
self.command_processor.commands)
self.speech_to_text = SpeechToText(
config.speechkit.key, "", config.speechkit.recognition.language)
self.text_to_speech = TextToSpeech(
config.speechkit.synthesis.cache_size, config.speechkit.key,
config.speechkit.synthesis.language,
config.speechkit.synthesis.speaker,
config.speechkit.synthesis.emotion,
config.speechkit.synthesis.speed)
self.record = SpeechCapture(config.record.silence_calculation_chunks,
config.record.speech_level_coefficient,
config.record.start_wait_chunks,
config.record.finish_wait_chunks)
def test_makeRequest(self):
face = ['will']
object = ['snowboard']
data = TwoFacesOneFakeOneReal()
results = [
"", "I can’t see any objects.", "No, I cannot see any snowboard.",
"I can see nico and will. But at least one of them is fake.",
"Yes, I am 84 percent sure that I can see will. Not the real one though. behind me on my left.",
"I don’t think anybody has a snowboard.", "",
"I cannot see any snowboard.",
"Yes, I am 84 percent sure that a fake will is here, behind me on my left."
]
for id in range(1, 9):
if id == 6:
continue
data.instruction_id = id
text_to_speech = TextToSpeech(data,
'en-EN',
input_faces=face,
input_objects=object)
out = text_to_speech.callback()
self.assertEqual(out, results[id])
def predict(speech_recognition=False, speech_synthesis=False):
''' Работа с обученной моделью seq2seq.
1. speech_recognition - включение распознавания речи с микрофона с помощью PocketSphinx
2. speech_synthesis - включение озвучивания ответов с помощью RHVoice '''
name_dataset = configure_file_names()
ttt = TextToText(f_name_w2v_model=f_name_w2v_model,
f_name_model=f_name_model,
f_name_model_weights=f_name_model_weights)
if speech_recognition:
print('[i] Загрузка языковой модели для распознавания речи...')
stt = SpeechToText('from_microphone', name_dataset)
if speech_synthesis:
print('[i] Загрузка синтезатора речи...')
tts = TextToSpeech('anna')
print()
question = ''
while (True):
if speech_recognition:
print('Слушаю...')
question = stt.get()
os.write(sys.stdout.fileno(), curses.tigetstr('cuu1'))
print('Вы: ' + question)
else:
question = input('Вы: ')
answer, lost_words = ttt.predict(question, True)
print('\t=> %s' % answer)
if len(lost_words) > 0:
print('[w] Потерянные слова: ' + ', '.join(lost_words) + '\n')
else:
print()
if speech_synthesis:
tts.get(answer)
def test_makeRequest(self):
face = ['will']
object = ['snowboard']
data = OneFaceReal()
results = [
"", "I can’t see any objects.", "No, I cannot see any snowboard.",
"I am 100 percent sure that I can see will behind me on my left.",
"Yes, I am 100 percent sure that I can see will behind me on my left.",
"I don’t think anybody has a snowboard.", "",
"I cannot see any snowboard.",
"I am 100 percent sure that will is here behind me on my left."
]
for id in range(1, 9):
if id == 6:
continue
data.instruction_id = id
text_to_speech = TextToSpeech(data,
'en-EN',
input_faces=face,
input_objects=object)
out = text_to_speech.callback()
self.assertEqual(out, results[id])
def test_makeRequest(self):
face = ['will']
object = ['snowboard']
data = TwoFacesSameReal()
results = [
"", "I can’t see any objects.", "No, I cannot see any snowboard.",
"I can see will and will.",
"Yes, I can actually see will more than once. I knew I shouldn’t have had that last pint last night. ",
"I don’t think anybody has a snowboard.", "",
"I cannot see any snowboard.",
"Yes, I can actually see will more than once. I knew I shouldn’t have had that last pint last night. One of them is here. "
]
for id in range(1, 9):
if id == 6:
continue
data.instruction_id = id
text_to_speech = TextToSpeech(data,
'en-EN',
input_faces=face,
input_objects=object)
out = text_to_speech.callback()
self.assertEqual(out, results[id])
from text_to_speech import TextToSpeech obj = TextToSpeech() obj.text_input() obj.create_gtts_object() obj.speech_save() obj.run()
def __init__(self):
self.queue = Queue()
self.alarm_running_queue = Queue()
self.media = MultimmediaController()
self.sched = Scheduler()
self.tts = TextToSpeech()
self.schedule = {}
#self.gcal = GoogleCalendar()
self.arduino = ArduinoController(0x08, self.queue)
self.gpio = RPiGPIO()
self.radio_db = RadioDB(
"/home/pi/Python/NetRadioAlarmClockPi/radio-settings.db")
self.webserver = webserver
self.webserver.add_station_cb = self.add_station
self.webserver.get_stations_cb = self.radio_db.get_stations
self.webserver.set_current_station_cb = self.set_current_station
self.webserver.get_current_station_cb = self.get_current_station
self.webserver.update_station_name_cb = self.radio_db.update_station_title
self.webserver.get_stream_playing_cb = self.media.get_playing
self.webserver.play_stream_cb = self.start_stream
self.webserver.stop_stream_cb = self.stop_stream
self.webserver.set_volume_cb = self.media.set_volume
self.webserver.get_schedule_cb = self.get_schedule
self.webserver.save_schedule_cb = self.save_schedule
self.webserver.alarm_running_queue = self.alarm_running_queue
self.alarm_running = False
self.update_interval = 60 * 5
self.state = "idle"
self.gpio.set_snooze_btn_callback(self.alarm_snooze_event)
self.webdev = False
self.stream_playing = False
self.alarm_running_queue.put(False)
self.last_snooze_time = None
self.button_press_count = 0
self.debouncing = False
# Set current station
url, station_title = self.get_current_station()
self.media.set_stream_url(url)
# Load radio alarm schedule and set up events if there are any
self.load_schedule()
self.create_events()
# Test event
test_time = datetime.datetime(2021, 2, 23, 20, 14, 0)
self.sched.add_event(test_time, self.alarm_event)
if len(sys.argv) > 1:
#print(sys.argv)
if sys.argv[1].replace("\r", "") == "webdev":
print("Web development mode")
self.webdev = True
if not self.webdev:
print("Resetting Arduino")
self.gpio.reset_arduino()
print("Done")
self.arduino.set_vol_change_callback(self.media.set_volume)
self.arduino.start_rot_enc_thread()
def main():
rospy.init_node("cyborg_audio")
playback = Playback()
text_to_speech = TextToSpeech()
rospy.spin()
def run(host, port, wsgi=False, https_mode=False):
''' Автовыбор доступного порта (если указан порт 0), загрузка языковой модели и нейронной сети и запуск сервера.
1. wsgi - True: запуск WSGI сервера, False: запуск тестового Flask сервера
2. https - True: запуск в режиме https (сертификат и ключ должны быть в cert.pem и key.pem), False: запуск в режиме http
Самоподписанный сертификат можно получить, выполнив: openssl req -x509 -newkey rsa:4096 -nodes -out temp/cert.pem -keyout temp/key.pem -days 365 '''
if port == 0: # Если был введён порт 0, то автовыбор любого доступного порта
try:
sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
sock.bind((host, 0))
port = sock.getsockname()[1]
log('выбран порт ' + str(port))
sock.close()
except socket.gaierror:
log('адрес ' + host + ':' + str(port) + ' некорректен',
level='error')
sock.close()
return
except OSError:
log('адрес ' + host + ':' + str(port) + ' недоступен',
level='error')
sock.close()
return
log('Flask v.' + flask_version + ', WSGIServer v.' + wsgi_version)
log('установлен максимальный размер принимаемых данных: {:.2f} Кб'.format(
max_content_length / 1024))
name_dataset = f_name_w2v_model_plays[
f_name_w2v_model_plays.rfind('w2v_model_') +
len('w2v_model_'):f_name_w2v_model_plays.rfind('.bin')]
log('загрузка обученной на наборе данных ' + name_dataset +
' модели seq2seq...')
global ttt
print()
ttt = TextToText(f_name_w2v_model=f_name_w2v_model_plays,
f_name_model=f_name_model_plays,
f_name_model_weights=f_name_model_weights_plays)
print()
log('загрузка языковой модели для распознавания речи...')
global stt
stt = SpeechToText('from_file', name_dataset)
log('загрузка синтезатора речи...')
global tts
tts = TextToSpeech('anna')
if wsgi:
global http_server
if https_mode:
log('WSGI сервер запущен на https://' + host + ':' + str(port) +
' (нажмите Ctrl+C или Ctrl+Z для выхода)')
else:
log('WSGI сервер запущен на http://' + host + ':' + str(port) +
' (нажмите Ctrl+C или Ctrl+Z для выхода)')
try:
if https_mode:
http_server = WSGIServer((host, port),
app,
log=app.logger,
error_log=app.logger,
keyfile='temp/key.pem',
certfile='temp/cert.pem')
else:
http_server = WSGIServer((host, port),
app,
log=app.logger,
error_log=app.logger)
http_server.serve_forever()
except OSError:
print()
log('адрес ' + host + ':' + str(port) + ' недоступен',
level='error')
else:
log('запуск тестового Flask сервера...')
try:
if https_mode:
app.run(host=host,
port=port,
ssl_context=('temp/cert.pem', 'temp/key.pem'),
threaded=True,
debug=False)
else:
app.run(host=host, port=port, threaded=True, debug=False)
except OSError:
print()
log('адрес ' + host + ':' + str(port) + ' недоступен',
level='error')
print("language code is %s"%(language_to_be_learnt))
print("How do you want to learn today?")
print(" ")
selector=input(("Enter 1 for translation help. Enter 2 to see the most used words translated for you. Enter 3 for a fun quiz. Enter 4 to hear %s pronunciation of specific words: ")%(language_full_name))
if(selector=="1"):
fine_selector=input(("Enter 0 to translate from english to %s. Enter 1 to translate from %s to english: ")%(language_full_name,language_full_name))
if(fine_selector=="0"):
times=int(input("Number of sentences to be translated: "))
for i in range(times):
text=(input(("Enter text to be translated to %s: ")%(language_full_name)))
translation =translate_text_to_foreign(text,language_to_be_learnt)
print("The translation is %s"%(translation))
speech_object = TextToSpeech(translation,language_to_be_learnt)
speech_object.speech()
print(" ")
if(i==times-1):
print("Thanks!!")
else:
times=int(input("Number of sentences to be translated: "))
for i in range(times):
text=(input("Enter text to be translated to english: "))
#language_to_be_learnt="English"
translation =translate_text_to_english(text,language_to_be_learnt)
print("The translation is %s"%(translation))
speech_object = TextToSpeech(translation,"en")
def main():
CHUNK = 1024
CHANNELS = 1
RATE = 16000
FORMAT = 'S16LE'
BYTE_RATE = 32000
STATUS_KWS = 0
STATUS_ASR = 1
STATUS_TTS = 2
IP = '203.113.152.90'
IP = '10.30.154.10'
URI = 'ws://{}:8892/client/ws/speech'.format(IP)
content_type = "audio/x-raw, layout=(string)interleaved, rate=(int){}, format=(string){}, channels=(int){}".format(
RATE, FORMAT, CHANNELS)
IS_ACTIVE = False
p = pyaudio.PyAudio()
stream = p.open(format=pyaudio.paInt16,
channels=CHANNELS,
rate=RATE,
input=True,
frames_per_buffer=CHUNK)
sr = None
sentence = None
kws = KeywordSpotting(graph='graph/conv_kws.pb',
labels='graph/conv_kws_labels.txt')
tts = TextToSpeech(url='http://{}/hmm-stream/syn'.format(IP),
voices='doanngocle.htsvoice')
brain = aiml.Kernel()
brain.learn("std-startup.xml")
brain.respond("load aiml b")
status = STATUS_KWS
print("===============================")
print("[INFO] Waiting keyword [xin chào | chào bot | hi bot] ...")
while True:
block = stream.read(CHUNK)
if status == STATUS_KWS:
if kws.spotting(block):
status = STATUS_ASR
print("[INFO] Keyword detected! Start recognize ...")
sr = SpeechRecognize(
url=URI + '?%s' %
(urllib.parse.urlencode([("content-type", content_type)])),
byterate=BYTE_RATE,
one_sentence=False)
sr_response = sr.get_response_queue()
elif status == STATUS_ASR:
if sr.is_alive():
sr.push_audio(block)
while sr_response.qsize() > 0:
msg = sr_response.get()
if msg == "EOS":
print("\rHuman: {}".format(sentence))
text = brain.respond(sentence)
audio = tts.get_speech(text)
status = STATUS_TTS
tts.play_audio(audio)
status = STATUS_ASR
if "tạm biệt" in sentence:
sr.close()
else:
sentence = msg
print("\r-----: {}".format(msg), end='')
else:
status = STATUS_KWS
print("===============================")
print("========= GOOD BYE!!! =========")
print("===============================")
print("[INFO] Waiting keyword ...")
elif status == STATUS_TTS:
if sr.is_alive():
sr.push_audio(bytearray(CHUNK))
time.sleep(1. * CHUNK / BYTE_RATE)
