def azure_verification(userId):
"""
record user audio and send request for identification
"""
try:
verification_filename = 'azure_verification.wav'
response_filename = 'ttsfiles/verification_response.mp3'
audio_length = 15
path = '/azure_verify'
params = {
'filename': verification_filename,
'userId': userId
}
os.system(countdown_cmd)
record.record(audio_length, verification_filename)
verification_response = file_request(path, verification_filename, params)
verification_response = json.loads(verification_response)
responseCode = verification_response.get('responseCode')
# provide audio feedback to user
if responseCode = 'SUCC':
confidence = verification_response.get('confidence')
st = "successfully verified user with confidence: {0}".format(confidence)
else:
def voiceit_enrollment(uid, phrases):
"""
takes userId and phrases, records user saying the phrase
and sends request to server with audio file and userId
before each recording, the program will do a countdown using text to speech
after receiveng server response, it converts it to audio to provide feedback to user
"""
try:
iterations = 3
audio_length = 5
enrollment_filename = 'enrollment.wav'
response_filename = 'ttsfiles/voiceit_enrollment_response.mp3'
print("supported phrases: \n", pretty_list(phrases))
count = 1
path = '/voiceit_enrollment'
filename = enrollment_filename
method = 'POST'
# 3 enrollments required for this API, keep trying until 3 successful enrollments are made
while count <= iterations:
# takes the audio file passed to it as argument and plays it for voice interaction
os.system(countdown_cmd)
record.record(audio_length, enrollment_filename)
phrase = file_to_text(enrollment_filename)
params = {
'phrase': phrase,
'filename': filename,
'userId': uid
}
#send request to local server (middleware)
enrollment_response = file_request("/voiceit_enrollment", filename, params = params)
enrollment_response = json.loads(enrollment_response)
enrollment_response_code = enrollment_response.get('responseCode')
# if enrollment attempt is successful, provide user feedback and increment count
if enrollment_response_code == 'SUCC':
count += 1
text = "attempt {0} of enrollment successful".format(count)
tts(text, response_filename)
else:
tts(enrolmment_response['message'], response_filename)
enrollment_cmd = 'mpg321 {}'.format(response_filename)
os.system(enrollment_cmd)
# delete audio files created every time
#if not explicitly deleted, they can cause problems for future requests
os.remove(enrollment_filename)
os.remove(response_filename)
os.system("mpg321 ttsfiles/EnrollmentSuccess.mp3")
return enrollment_response
except Exception as e:
print(str(e))
return str(e)
def voiceit_enrollment(uid, phrases):
try:
iterations = 3
audio_length = 5
enrollment_filename = 'enrollment.wav'
response_filename = 'voiceit_enrollment_response'
print("supported phrases: \n", pretty_list(phrases))
count = 1
path = '/voiceit_enrollment'
filename = enrollment_filename
method = 'POST'
# 3 enrollments required for this API, keep trying until 3 successful enrollments are made
while count <= iterations:
os.system("mpg321 ttsfiles/countdown.mp3")
record.record(audio_length, enrollment_filename)
phrase = file_to_text(enrollment_filename)
params = {
'phrase': phrase,
'filename': filename,
'userId': uid
}
#send request to local server (middleware)
enrollment_response = file_request("/voiceit_enrollment", filename, params = params)
enrollment_response_code = enrollment_response.get('responseCode')
# if enrollment attempt is successful, provide user feedback and increment count
if enrollment_response_code == 'SUCC':
count += 1
text = "attempt {0} of enrollment successful".format(count)
tts(text, response_filename)
else:
tts(r['message'], response_filename)
enrollment_cmd = 'mpg321 ttsfiles/{}.mp3'.format(response_filename)
os.system(enrollment_cmd)
os.system("mpg321 ttsfiles/EnrollmentSuccess.mp3")
return enrollment_response
except Exception as e:
print(str(e))
return str(e)
def azure_verification_enrollment(userId, phrases):
"""
given an user Id, record the audio file and send it to the server along with the Id
"""
try:
enrollment_filename = 'verification_enrollment.wav'
response_filename = 'ttsfiles/verification_response.mp3'
iterations = 3
audio_length = 15
path = '/azure_verification_enrollment'
params = {
'userId': userId,
'filename': enrollment_filename
}
print("phrases: ", pretty_list(phrase))
count = 1
# 3 registrations needed, keep trying until 3 successfull attemps have been made
while count <= iterations:
os.system(countdown_cmd)
record.record(audio_length, enrollment_filename)
enrollment_response = file_request(path, enrollment_filename, params)
responseCode = enrollment_response.get('responseCode')
# provide audio feedback to the user
if responseCode == 'SUCC':
st = "attempt {} of azure verification enrollment successfull".format(count)
tts(st, response_filename)
count += 1
else:
tts("error trying to enroll user for verification ", response_filename)
speech_cmd = 'mpg321 {}'.format(response_filename)
os.system(speech_cmd)
os.remove(enrollment_filename)
os.remove(response_filename)
os.system("mpg321 ttsfiles/EnrollmentSuccess.mp3")
return enrollment_response
except Exception as e:
print(str(e))
return str(e)
def azure_identification():
"""
record user audio and send request for identification
"""
try:
identification_filename = 'azure_identification.wav'
response_filename = 'ttsfiles/identification_response.mp3'
audio_length = 15
path = '/azure_identification'
params = {
'filename': identification_filename,
}
os.system(countdown_cmd)
record.record(audio_length, identification_filename)
identification_response = file_request(path, identification_filename, params)
identification_response = json.loads(identification_response)
responseCode = idenfication_response.get('identification_response')
# provide audio feedback to user
if responseCode == 'SUCC':
name = idenfication_response.get('username')
if name:
st = "First step completed, {0} was successfully identified. Please wait for the complete verification".format(name)
else:
st = "First step completed, user was successfully identified. Please wait for the complete verification"
else:
st = "Error identifying user"
tts(st, response_filename)
identification_cmd = 'mpg321 {0}'.format(response_filename)
os.system(identification_cmd)
os.remove(identification_filename)
os.remove(response_filename)
return identification_response
except Exception as e:
print(str(e))
return str(e)
def azure_identification_enrollment(userId):
"""
given an user Id, record the audio file and send it to the server along with the Id
"""
try:
enrollment_filename = 'identification_enrollment.wav'
response_filename = 'ttsfiles/idenfication_response.mp3'
audio_length = 30
path = "/azure_identification_enrollment"
params = {
'userId': userId,
'filename': enrollment_filename
}
# do countdown before starting to record and then send request with audio file
os.system(countdown_cmd)
record.record(audio_length, enrollment_filename)
identification_response = file_request(path, enrollment_filename, params)
identification_response = json.loads(identification_response)
responseCode = identification_response.get('responseCode')
# provide audio feedback to the user
if responseCode == 'SUCC':
tts('successfully enrolled user for identification', response_filename)
else:
message = idenfication_response.get('message')
if message:
tts(message, response_filename)
else:
tts('error trying to enroll user for identification', response_filename)
speech_cmd = 'mpg321 {}'.format(response_filename)
os.system(speech_cmd)
os.remove(response_filename)
os.remove(enrollment_filename)
return identification_response
except Exception as e:
print(str(e))
return str(e)
def voiceit_verification(groupId):
"""
takes the voiceit group id, records audio to be used for identification & verification
first performs the identification, if successful it proceeds to verification
if any of the 2 fails, it just returns error message
"""
try:
# identification process
verification_filename = 'verify.wav'
identification_response_filename = 'ttsfiles/identification_response.mp3'
audio_length = 5
verification_response_filename = 'ttsfiles/verification_response.mp3'
# do countdown and record
os.system("mpg321 ttsfiles/startedVerification.mp3")
os.system(countdown_cmd)
record.record(audio_length, verification_filename)
phrase = file_to_text(verification_filename)
params = {
'phrase': phrase,
'filename': verification_filename,
'groupId': groupId
}
# send request to server, with audio file and params as HTTP headers
identification_response = file_request('/voiceit_identify', verification_filename, params)
identification_response = json.loads(identification_response)
responseCode = identification_response.get('responseCode')
# if identification is successful proceed to verification using the recorded file
# and the userId returned in the response to identification
if responseCode == 'SUCC':
uid = identification_response.get('userId')
name = identification_response.get('username')
params = {
'phrase': phrase,
'filename': verification_filename,
'userId': uid
}
# if the name of the user is available, use that name. Otherwise, just say user
if name:
s = "First step completed, {0} was successfully identified. Please wait for the complete verification".format(name)
else:
s = "First step completed, user was successfully identified. Please wait for the complete verification"
# provide feedback of identification request
tts(s, identification_response_filename)
identification_cmd = 'mpg321 {}'.format(identification_response_filename)
os.system(identification_cmd)
# send verification request
verification_response = file_request('/voiceit_verify', verification_filename, params)
verification_response = json.loads(verification_response)
verification_response_code = verification_response.get('responseCode')
# if request is successful, get the confidence
if verification_response_code == 'SUCC':
confidence = verification_response.get('confidence')
st = "successfully verified user with confidence: {0}".format(confidence)
else:
st = verification_response.get('message')
# provide feeback of verification request
tts(st, verification_response_filename)
verification_cmd = 'mpg321 {}'.format(verification_response_filename)
os.system(verification_cmd)
# explicitly remove all audio files used for voice interaction
os.remove(identification_response_filename)
os.remove(verification_response_filename)
os.remove(verification_filename)
return verification_response
else:
# error message of identification
st = identification_response.get("message")
tts(st, identification_response_filename)
identification_cmd = 'mpg321 ttsfiles/{}.mp3'.format(identification_response_filename)
os.system(identification_cmd)
os.remove(identification_response_filename)
os.remove(verification_response_filename)
os.remove(verification_filename)
return identification_response
except Exception as e:
print(str(e))
return str(e)
def voiceit_verification(groupId):
try:
# identification process
verification_filename = 'verify.wav'
identification_response_filename = 'identification_response'
audio_length = 5
verification_response_filename = 'verification_response'
os.system("mpg321 ttsfiles/startedVerification.mp3")
os.system("mpg321 ttsfiles/countdown.mp3")
record.record(audio_length, verification_filename)
phrase = file_to_text(verification_filename)
params = {
'phrase': phrase,
'filename': verification_filename,
'groupId': groupId
}
identification_response = file_request('/voiceit_identification', verification_filename, params)
responseCode = res.get('responseCode')
# if identification is successful proceed to verification using the recorded file
# and the userId returned in the response to identification
if responseCode == 'SUCC':
print(res)
uid = res.get('userId')
name = res.get('username')
# if the name of the user is available, use that name. Otherwise, just say user
if name:
s = "First step completed, {0} was successfully identified. Please wait for the complete verification".format(name)
else:
s = "First step completed, user was successfully identified. Please wait for the complete verification"
tts(s, identification_response_filename)
identification_cmd = 'mpg321 ttsfiles/{}.mp3'.format(identification_response_filename)
os.system(identification_cmd)
params = {
'phrase': phrase,
'filename': verification_filename,
'userId': uid
}
verification_response = file_request('/voiceit_verification', verification_filename, params)
verification_response_code = verification_response.get('responseCode')
if verification_response_code == 'SUCC':
confidence = verification_response.get('confidence')
st = "successfully verified user with confidence: {0}".format(confidence)
else:
st = verification_response.get('message')
print(verification_response)
tts(st, verification_response_filename)
verification_cmd = 'mpg321 ttsfiles/{}.mp3'.format(verification_response_filename)
os.system(verification_cmd)
return verification_response
else:
st = identification_response.get("message")
tts(st, identification_response_filename)
identification_cmd = 'mpg321 ttsfiles/{}.mp3'.format(identification_response_filename)
os.system(identification_cmd)
return identification_response
except Exception as e:
print(str(e))
return str(e)