started doing weird stuff for speech recognition

This commit is contained in:
Pavel Lutskov
2019-01-30 17:37:26 +01:00
parent f050c3464a
commit e368cd9abe
7 changed files with 111 additions and 4 deletions

View File

@@ -11,6 +11,7 @@ find_package(catkin REQUIRED COMPONENTS
tf tf
aruco aruco
message_generation message_generation
actionlib_msgs
) )
add_service_files( add_service_files(
@@ -19,9 +20,16 @@ add_service_files(
InformController.srv InformController.srv
) )
generate_messages(DEPENDENCIES std_msgs) add_action_files(
DIRECTORY action
FILES
RequestSpeech.action
)
generate_messages(DEPENDENCIES actionlib_msgs std_msgs)
catkin_package( catkin_package(
CATKIN_DEPENDS message_runtime CATKIN_DEPENDS message_runtime actionlib_msgs
INCLUDE_DIRS include INCLUDE_DIRS include
) )

View File

@@ -0,0 +1,5 @@
string[] vocabulary
----
string result
----
bool success

View File

@@ -48,6 +48,7 @@
<build_depend>std_msgs</build_depend> <build_depend>std_msgs</build_depend>
<build_depend>aruco</build_depend> <build_depend>aruco</build_depend>
<build_depend>message_generation</build_depend> <build_depend>message_generation</build_depend>
<build_depend>actionlib_msgs</build_depend>
<run_depend>cv_bridge</run_depend> <run_depend>cv_bridge</run_depend>
<run_depend>image_transport</run_depend> <run_depend>image_transport</run_depend>
<run_depend>roscpp</run_depend> <run_depend>roscpp</run_depend>
@@ -56,6 +57,7 @@
<run_depend>std_msgs</run_depend> <run_depend>std_msgs</run_depend>
<run_depend>aruco</run_depend> <run_depend>aruco</run_depend>
<run_depend>message_runtime</run_depend> <run_depend>message_runtime</run_depend>
<run_depend>actionlib_msgs</run_depend>
<!-- The export tag contains other, unspecified, tags --> <!-- The export tag contains other, unspecified, tags -->

11
script/calibrator.py Executable file
View File

@@ -0,0 +1,11 @@
#! /usr/bin/env python
import rospy
from controller import inform_controller_factory
_inform_controller = inform_controller_factory('calibrator')
if __name__ == '__main__':
rospy.init_node('teleoperation/calibrator')

View File

@@ -73,6 +73,6 @@ def handle_request(r):
if __name__ == '__main__': if __name__ == '__main__':
rospy.init_node('controller', log_level=rospy.INFO) rospy.init_node('teleoperation/controller', log_level=rospy.INFO)
ic = rospy.Service('inform_controller', InformController, handle_request) ic = rospy.Service('inform_controller', InformController, handle_request)
rospy.spin() rospy.spin()

81
script/speech.py Executable file
View File

@@ -0,0 +1,81 @@
#! /usr/bin/env python
import os
import rospy
import actionlib
from naoqi import ALProxy, ALModule, ALBroker
speech_broker = None
almem = None
sas = actionlib.SimpleActionServer()
def request_speech(goal):
recognized = ''
if not speech_detector.start_speech(goal):
sas.publish_feedback(False)
return
while not sas.is_preempt_requested() and not speech_detector.have_word():
rospy.Rate(10).sleep()
speech_detector.stop_speech(goal)
if speech_detector.have_word():
recognized = speech_detector.get_recognized_and_erase()
sas.publish_feedback(True)
sas.set_succeeded(recognized)
class SpeechDetectorModule(ALModule):
def __init__(self, name):
ALModule.__init__(self, name)
self.recognized = None
self.subid = 'teleoperation_speech'
self.asr = ALProxy('ALSpeechRecognition')
self.asr.setLanguage('English')
self.stop_speech()
def start_speech(self, voc):
if self.running:
return False
# self.stop_speech()
self.asr.setVocabulary(voc, False)
self.asr.subscribe(self.subid)
self.asr.pause(False)
self.running = True
return True
def have_word(self):
return self.recognized is not None
def get_recognized_and_erase(self):
result = self.recognized
self.recognized = None
return result
def stop_speech(self):
self.asr.pause(True)
self.asr.unsubscribe(self.subid)
self.running = False
def on_word_recognized(self, *_args):
rospy.loginfo(almem.getData('WordRecognized'))
if __name__ == '__main__':
rospy.init_node('speech')
speech_broker = ALBroker('speech_broker', '0.0.0.0', 0,
os.environ['NAO_IP'], 9559)
speech_detector = SpeechDetectorModule('speech_detector')
almem = ALProxy('ALMemory')
almem.subscribeToEvent("WordRecognized",
"speech_detector",
"on_word_recognized")
speech_detector.start_speech(['start', 'stop', 'pause'])
while not rospy.is_shutdown():
rospy.Rate(4).sleep()
speech_detector.stop_speech()
speech_broker.shutdown()

View File

@@ -37,7 +37,7 @@ def _speed(pos, interval):
if __name__ == '__main__': if __name__ == '__main__':
rospy.init_node('walker') rospy.init_node('teleoperation/walker')
rospy.wait_for_service('inform_controller') rospy.wait_for_service('inform_controller')
ll = tf.TransformListener() ll = tf.TransformListener()
mp = ALProxy('ALMotion', os.environ['NAO_IP'], 9559) mp = ALProxy('ALMotion', os.environ['NAO_IP'], 9559)