from Recorder import Recorder
from SpeechToText import speech_to_text
from TextTranslator import translate_text
recorder = Recorder('/tmp/out.wav')
recorder.start()
print ">>> Recording"
raw_input('press enter to stop ...')
recorder.stop()
print "<<< Stopped "
text = speech_to_text('es-MX', '/tmp/out.wav')
print text
translation = translate_text('en', text)
print translation
def myexit():
g_recorder.stop()
exit()
if __name__ == '__main__':
gpu_setting()
model = load_model('md-normal-lr2.h5')
if os.path.exists('Recorder-tmpwav.wav'):
os.remove('Recorder-tmpwav.wav')
rd = Recorder()
rd.start()
g_recorder = rd
signal.signal(signal.SIGINT, myexit)
signal.signal(signal.SIGTERM, myexit)
#time.sleep(12)
while True:
try:
key = input('\r\nPlease input, Enter:detect p:Predict k:Stop (Enter/p/k)?')
except Exception as e:
continue
if key == 'k':
myexit()
class Mouvements(object):
def __init__(self):
IP, PORT = "127.0.0.1", 9559
self.vx = 0.0
self.vy = 0.0
self.vth = 0.0
self.dt = 0.227
self.joints = []
self.io_file = IO_file()
self.is_recording = False
self.postureProxy = ALProxy("ALRobotPosture", IP, PORT)
self.motionProxy = ALProxy("ALMotion", IP, PORT)
self.motionProxy.setWalkArmsEnabled(True, True)
self.motionProxy.setMotionConfig([["ENABLE_FOOT_CONTACT_PROTECTION", False]])
self.motionProxy.setFallManagerEnabled(False)
def stiffness(self, go_rigide):
parametre = 1.0 if go_rigide else 0.0
self.motionProxy.stiffnessInterpolation("Body", parametre, 1.0)
def go_debout(self):
self.stiffness(True)
self.postureProxy.goToPosture("Stand", 0.8)
def go_assis(self):
self.stiffness(True)
self.postureProxy.goToPosture("Sit", 0.8)
self.stiffness(False)
def go_ventre(self):
self.stiffness(True)
self.postureProxy.goToPosture("LyingBelly", 0.8)
self.stiffness(False)
def go_dos(self):
self.stiffness(True)
self.postureProxy.goToPosture("LyingBack", 0.8)
self.stiffness(False)
def go_dos(self):
self.stiffness(True)
self.postureProxy.goToPosture("Crouch", 0.8)
self.stiffness(False)
def go_move(self):
self.motionProxy.setWalkTargetVelocity(self.vx, self.vy, self.vth, 0.8)
def set_vx(self, vx): self.vx = vx
def set_vy(self, vy): self.vy = vy
def set_vth(self, vth): self.vth = vth
def update_joints(self): self.joints = self.motionProxy.getAngles("Body", False)
def save_joints(self, write):
self.update_joints()
if write: fichier = self.io_file.init_write_joints()
self.io_file.add_joints(self.joints, self.dt)
if write: self.io_file.write_joints(fichier)
def record(self):
self.is_recording = not self.is_recording
if self.is_recording:
self.recorder = Recorder(self)
self.recorder.continuer = True
self.recorder.start()
else:
self.recorder.continuer = False
self.recorder.join()
def apply_joints_from_file(self, only_last):
temps, angles = self.io_file.read_joints(only_last)
if len(angles) == 0 : return
self.motionProxy.angleInterpolation("Body", angles, temps, True)
def move_head(self, vx, vy):
self.motionProxy.angleInterpolation(["HeadYaw", "HeadPitch"], [-0.1 * float(vx), -0.1 * float(vy)], [0.2, 0.2], False)
class Experiment:
def __init__(self):
self.name = datetime.now().strftime("%Y-%m-%d.%H_%M_%S.%f")
self.folder = os.path.abspath(os.curdir) + '\\' + self.name
self.recorder = Recorder('')
self.emotions = []
self.photos = []
self.brain_waves = []
self.res_dataset = []
def set_name(self, name):
self.name = name
self.folder = self.rsplit('\\', 1)[0] + name
def set_folder(self, folder):
self.folder = folder + '\\' + self.name
def start_record(self):
os.mkdir(
self.folder
) #надо удалять если ошибка или создавать только после сигнала об удаче
self.recorder.start([self.name, self.folder])
def stop_record(self):
self.recorder.stop()
self.save_results()
return True
def save_results(self):
self.prepare_brain_waves()
self.prepare_photos()
self.prepare_emotions()
self.make_dataset(self.match_data())
DataManager.write_dataset(
self.folder + '\\' + 'res_' + self.name + '.tsv', self.res_dataset)
def match_data(self):
count_waves = 1
count_photos = 1
dataset = []
while count_waves < len(self.brain_waves) and count_photos < len(
self.photos):
# узнать время первого снимка
time_begin = datetime.strptime(self.photos[count_photos - 1][-1],
"%Y-%m-%d.%H_%M_%S.%f")
# узнать время следующего снимка
time_end = datetime.strptime(self.photos[count_photos][-1],
"%Y-%m-%d.%H_%M_%S.%f")
# узнать время первого измерения мозговых волн
current_time = datetime.strptime(
self.brain_waves[count_waves - 1][-1], "%Y-%m-%d.%H_%M_%S.%f")
if current_time > time_end:
count_photos += 1
elif current_time < time_begin:
count_waves += 1
else:
element = self.brain_waves[count_waves - 1]
element.extend(self.photos[count_photos])
dataset.append(element)
#dataset.append([self.brain_waves[count_waves-1], self.photos[count_photos]])
count_waves += 1
return dataset
def prepare_photos(self):
photos = DataManager.read_file(self.folder + '\\' + 'photos_log.txt')
for photo in photos:
time = photo.rsplit('.', 1)[0]
time = time.rsplit('\\', 1)[-1]
self.photos.append([photo, time])
def prepare_brain_waves(self):
brain_waves_strings = DataManager.read_file(self.folder + '\\' +
self.name + '.tsv')
for line in brain_waves_strings:
self.brain_waves.append(line.rsplit('\t', -1))
self.brain_waves = self.brain_waves[1:]
return len(self.brain_waves) != 0
def prepare_emotions(self):
emo_analyzer = EmotionAnalyzer()
#self.emotions.append(['Angry', 'Disgust', 'Fear', 'Happy', 'Sad', 'Surprise', 'Neutral'], 'time')
for i in range(len(self.photos)):
faces = ImageProcessor.find_faces(
self.photos[i][0]) # что если несколько лиц?
if len(faces) == 0:
print('No face on ' + self.photos[i][0])
elif len(faces) > 1:
print('More than 1 face on ' + self.photos[i][0])
else:
element = []
element.extend(emo_analyzer.predict(faces[0]))
element.append(self.photos[i][-1])
self.emotions.append(element)
def make_dataset(self, dataset):
print(dataset[0][-1])
print(self.emotions[0][-1])
print((datetime.strptime(dataset[0][-1], "%Y-%m-%d.%H_%M_%S.%f") - \
datetime.strptime(self.emotions[0][-1], "%Y-%m-%d.%H_%M_%S.%f")).microseconds)
count_waves = 0
count_emotions = 0
while count_waves < len(dataset):
if (datetime.strptime(dataset[count_waves][-1], "%Y-%m-%d.%H_%M_%S.%f") - \
datetime.strptime(self.emotions[count_emotions][-1], "%Y-%m-%d.%H_%M_%S.%f")).microseconds == 0:
dataset[count_waves] = dataset[
count_waves][:len(dataset[count_waves]) - 2]
dataset[count_waves].extend(self.emotions[count_emotions])
self.res_dataset.append(dataset[count_waves])
elif datetime.strptime(dataset[count_waves][-1], "%Y-%m-%d.%H_%M_%S.%f") > \
datetime.strptime(self.emotions[count_emotions][-1], "%Y-%m-%d.%H_%M_%S.%f"):
count_emotions += 1
count_waves -= 1
count_waves += 1
