-

audio.py

@@ -11,7 +11,7 @@
 from pocketsphinx.pocketsphinx import *
 from sphinxbase.sphinxbase import *
 
-CHUNK_SIZE = 131072
+CHUNK_SIZE = 4096
 FORMAT = pyaudio.paInt16
 RATE = 44100
 
@@ -116,38 +116,42 @@ def speech_recog(self, model):
     # Create a decoder with certain model
     config = Decoder.default_config()
     config.set_string('-hmm', '/usr/local/share/pocketsphinx/model/en-us/en-us')
-    config.set_string('-kws', MODELDIR + model + '.txt')
-    #config.set_string('-lm', MODELDIR + model + '.lm')
+    config.set_int('-ds', 2)
+    config.set_int('-topn', 3)
+    config.set_int('-maxwpf', 5)
+    #config.set_string('-kws', MODELDIR + model + '.txt')
+    config.set_string('-lm', MODELDIR + model + '.lm')
     config.set_string('-dict', MODELDIR + model + '.dict')
     decoder = Decoder(config)
 
     p = pyaudio.PyAudio()
+    logging.info("device info: " + str(p.get_device_info_by_index(0)))
     #stream = p.open(format=pyaudio.paInt16, channels=1, input_device_index=0, rate=16000, input=True, frames_per_buffer=1024)
     stream = p.open(format=FORMAT, channels=1, input_device_index=0, rate=RATE,
                     input=True,
-                    frames_per_buffer=CHUNK_SIZE)
+                    frames_per_buffer=CHUNK_SIZE*256)
     stream.start_stream()
     decoder.start_utt()
     tstamp = time.time()
     recog_text = ''
 
-    while time.time() - tstamp < 10.0:
+    while len(recog_text) < 1:
       try:
         buf = stream.read(CHUNK_SIZE)
+        logging.info("actual voice")
+        decoder.process_raw(buf, False, False)
+        if decoder.hyp().hypstr != '':
+          recog_text += decoder.hyp().hypstr
+          print "text: " + decoder.hyp().hypstr
+          tstamp = time.time()
       except IOError as ex:
         if ex[1] != pyaudio.paInputOverflowed:
           raise
         buf = '\x00' * CHUNK_SIZE #white noise
         logging.info("white noise") 
-
-      decoder.process_raw(buf, False, False)
-      try:
-        if decoder.hyp().hypstr != '':
-          recog_text += decoder.hyp().hypstr
-          print "text: " + decoder.hyp().hypstr
-          tstamp = time.time()
       except AttributeError:
         pass
+
     decoder.end_utt()
     stream.close()
     p.terminate()

coderbot.cfg

@@ -1 +1 @@
-{"move_tr_speed": "70", "move_fw_elapse": "1.5", "camera_color_object_size_min": "20", "camera_path_object_size_min": "20", "load_at_start": "", "move_tr_elapse": "1", "sound_stop": "$shutdown.mp3", "show_control_move_commands": "true", "prog_level": "adv", "prog_scrollbars": "true", "move_fw_speed": "100", "camera_color_object_size_max": "8000", "sound_shutter": "$shutter.mp3", "show_page_prefs": "true", "ctrl_hud_image": "", "button_func": "none", "ctrl_fw_elapse": "-1", "ctrl_tr_elapse": "-1", "move_power_angle_2": "20", "move_power_angle_3": "20", "move_power_angle_1": "15", "show_page_program": "true", "sound_start": "$startup.mp3", "camera_exposure_mode": "auto", "ctrl_tr_speed": "80", "ctrl_fw_speed": "100", "camera_refresh_timeout": "0.1", "camera_jpeg_quality": "10", "prog_maxblocks": "-1", "move_motor_mode": "servo", "camera_path_object_size_max": "8000", "show_page_control": "true"}
+{"move_tr_speed": "70", "move_fw_elapse": "1.5", "camera_color_object_size_min": "20", "camera_path_object_size_min": "20", "load_at_start": "", "move_tr_elapse": "1", "sound_stop": "$shutdown.mp3", "show_control_move_commands": "true", "prog_level": "adv", "prog_scrollbars": "true", "move_fw_speed": "100", "camera_color_object_size_max": "8000", "sound_shutter": "$shutter.mp3", "show_page_prefs": "true", "ctrl_hud_image": "", "button_func": "none", "ctrl_fw_elapse": "-1", "ctrl_tr_elapse": "-1", "move_power_angle_2": "20", "move_power_angle_3": "20", "move_power_angle_1": "15", "move_motor_trim": "0.95", "show_page_program": "true", "sound_start": "$startup.mp3", "camera_exposure_mode": "auto", "ctrl_tr_speed": "80", "ctrl_fw_speed": "100", "camera_refresh_timeout": "0.1", "camera_jpeg_quality": "10", "prog_maxblocks": "-1", "move_motor_mode": "servo", "camera_path_object_size_max": "8000", "show_page_control": "true"}

coderbot.py

@@ -1,6 +1,7 @@
 import os
 import time
 import pigpio
+import config
 
 PIN_MOTOR_ENABLE = 22
 PIN_LEFT_FORWARD = 25
@@ -20,13 +21,14 @@ def coderbot_callback(gpio, level, tick):
 class CoderBot:
   _pin_out = [PIN_MOTOR_ENABLE, PIN_LEFT_FORWARD, PIN_RIGHT_FORWARD, PIN_LEFT_BACKWARD, PIN_RIGHT_BACKWARD, PIN_SERVO_3, PIN_SERVO_4]
 
-  def __init__(self, servo=False):
+  def __init__(self, servo=False, motor_trim_factor=1.0):
     self.pi = pigpio.pi('localhost')
     self.pi.set_mode(PIN_PUSHBUTTON, pigpio.INPUT)
     self._cb = dict()
     self._cb_last_tick = dict()
     self._cb_elapse = dict()
     self._servo = servo
+    self._motor_trim_factor = motor_trim_factor
     if self._servo:
       self.motor_control = self._servo_motor
     else:
@@ -42,16 +44,16 @@ def __init__(self, servo=False):
   the_bot = None
 
   @classmethod
-  def get_instance(cls, servo=False):
+  def get_instance(cls, servo=False, motor_trim_factor=1.0):
     if not cls.the_bot:
       cls.the_bot = CoderBot(servo)
     return cls.the_bot
 
   def move(self, speed=100, elapse=-1):
-    self.motor_control(speed_left=speed, speed_right=speed, elapse=elapse)
+    self.motor_control(speed_left=speed / self._motor_trim_factor, speed_right=speed * self._motor_trim_factor, elapse=elapse)
 
   def turn(self, speed=100, elapse=-1):
-    self.motor_control(speed_left=speed, speed_right=-speed, elapse=elapse)
+    self.motor_control(speed_left=speed / self._motor_trim_factor, speed_right=-(speed * self._motor_trim_factor), elapse=elapse)
 
   def forward(self, speed=100, elapse=-1):
     self.move(speed=speed, elapse=elapse)

main.py

@@ -216,7 +216,7 @@ def run_server():
   global audio
   try:
     app.bot_config = Config.read()
-    bot = CoderBot.get_instance(servo=(app.bot_config.get("move_motor_mode")=="servo"))
+    bot = CoderBot.get_instance(servo=(app.bot_config.get("move_motor_mode")=="servo"), motor_trim_factor = float(app.bot_config.get('move_motor_trim', '1.0')))
     cam = Camera.get_instance()
     motion = Motion.get_instance()
     audio = Audio.get_instance()

psmodels/model_simple.txt

@@ -1,3 +1,4 @@
+CODERBOT /1e-10/
 MOVE /1e-10/
 FORWARD /1e-10/
 BACKWARD /1e-10/

templates/config.html

@@ -115,6 +115,8 @@ <h1>CoderBot</h1>
     <div id="t_movement" class="ui-body-d ui-content">
      <label for="i_move_motor_mode">{% trans %}Motor control mode{% endtrans %}</label>
      <input type="text" id="i_move_motor_mode" name="move_motor_mode" value="{{config.move_motor_mode}}">
+     <label for="i_move_motor_trim">{% trans %}Trim factor (1.0 = center){% endtrans %}</label>
+     <input type="text" id="i_move_motor_trim" name="move_motor_trim" value="{{config.move_motor_trim}}">
      <label for="i_move_power_angle_1">{% trans %}Power (target angle -15){% endtrans %}</label>
      <input type="text" id="i_move_power_angle_1" name="move_power_angle_1" value="{{config.move_power_angle_1}}"> 
      <label for="i_move_power_angle_2">{% trans %}Power (target angle -4){% endtrans %}</label>