def test_one_char_phonemes(self):
english_sentence_list = "./pizzeria.sentences.txt"
new_lm_folder = "./pizzeria"
#generate n-gram files
from docks.docks import ngram_generation
ngram_generation.generate_ngram_language_model(
training_sentences_file=english_sentence_list,
n_gram_order=2,
new_lm_folder=new_lm_folder)
from docks.docks.ngram_postprocessor import NgramPostprocessor
ngram_postprocessor = NgramPostprocessor(lm_folder_path=new_lm_folder,
phonemes=True)
#ngram initializiation is non blocking so wait some time
#import time
#time.sleep(5)
test_sentence = [
"B", "R", "IH", "NG", " ", "M", "IY", " ", "AH", " ", "HH", "AE",
"M", " ", "P", "IY", "T", "Z", "AH"
]
from asr_nlp_tools.utils.phoneme_converter import convert_arpabet_to_one_char
test_sentence = convert_arpabet_to_one_char(test_sentence)
hypothesis = ngram_postprocessor.recognize(test_sentence)
self.assertEqual(hypothesis, "bring me a ham pizza")
def atest_string_input_to_phoneme_model(self):
english_sentence_list = "./pizzeria.sentences.txt"
new_lm_folder = "./pizzeria"
#generate n-gram files
from docks.docks import ngram_generation
ngram_generation.generate_ngram_language_model(
training_sentences_file=english_sentence_list,
n_gram_order=2,
new_lm_folder=new_lm_folder)
from docks.docks.ngram_postprocessor import NgramPostprocessor
ngram_postprocessor = NgramPostprocessor(lm_folder_path=new_lm_folder,
phonemes=True)
#ngram initializiation is non blocking so wait some time
#import time
#time.sleep(5)
test_sentence = "pleas bring mee a han pizza"
self.assertRaises(TypeError, ngram_postprocessor.recognize,
test_sentence)
def atest_string(self):
english_sentence_list = "./pizzeria.sentences.txt"
new_lm_folder = "./pizzeria"
#generate n-gram files
from docks.docks import ngram_generation
ngram_generation.generate_ngram_language_model(
training_sentences_file=english_sentence_list,
n_gram_order=2,
new_lm_folder=new_lm_folder)
from docks.docks.ngram_postprocessor import NgramPostprocessor
ngram_postprocessor = NgramPostprocessor(lm_folder_path=new_lm_folder,
phonemes=False)
#ngram initializiation is non blocking so wait some time
#import time
#time.sleep(5)
test_sentence = "pleas bring mee a han pizza"
hypothesis = ngram_postprocessor.recognize(test_sentence)
self.assertEqual(hypothesis, "please bring me a ham pizza")
from docks.docks import suppress_alsa_warnings
import speech_recognition as sr
from pprint import pprint
if __name__ == "__main__":
os.system('tput reset')
terminal_name = 'DOCKS EXAMPLE'
sys.stdout.write("\x1b]2;" + terminal_name + "\x07")
english_sentence_list = "./pizzeria.sentences.txt"
new_lm_folder = "./pizzeria"
#generate n-gram files
from docks.docks import ngram_generation
ngram_generation.generate_ngram_language_model(
training_sentences_file=english_sentence_list,
n_gram_order=3,
new_lm_folder=new_lm_folder)
from docks.docks.ngram_postprocessor import NgramPostprocessor
ngram_postprocessor = NgramPostprocessor(lm_folder_path=new_lm_folder)
#ngram initializiation is non blocking so wait some time
import time
time.sleep(5)
# obtain audio from the microphone
google_recognizer = sr.Recognizer()
#adjust the microphone to background noise
with sr.Microphone() as source:
google_recognizer.adjust_for_ambient_noise(source)