How to use the mtcnn.mtcnn.MTCNN function in mtcnn

def detect_face(img_path, detector=MTCNN()):
    """
    detect face with MTCNN
    :param img_path:
    :return:
    """
    img = cv2.imread(img_path)
    if detector is None:
        detector = MTCNN()
    mtcnn_result = detector.detect_faces(img)

    return mtcnn_result

def __init__(self, preprocessed_image_size=128):
        from mtcnn.mtcnn import MTCNN
        self.database_directory = '../LAP Apparent Age V2'
        self.face_detector = MTCNN(steps_threshold=[0.5, 0.6, 0.6])
        self.preprocessed_image_size = preprocessed_image_size

def face_detection(username):
        x = 0
        ''' Get user media and scan it for a face'''
        user_id = bot.get_user_id_from_username(username)
        medias = bot.get_user_medias(user_id, filtration=False)
        for media in medias:
            while x < 1:
                try:
                    bot.logger.info(media)
                    path = bot.download_photo(media, folder=username)
                    img = cv2.imread(path)
                    detector = MTCNN()
                    detect = detector.detect_faces(img)
                    if not detect:
                        Bots.save_user_info(ig_username, "no face detected " + bot.get_link_from_media_id(media))
                        bot.logger.info("save user info")
                        bot.logger.info("no face detected " + bot.get_link_from_media_id(media))
                        x += 1

                    elif detect:
                        Bots.save_user_info(ig_username, "there was a face detected")
                        bot.logger.info("save user info")
                        bot.logger.info("there was a face detected")
                        bot.api.like(media)
                        display_url = bot.get_link_from_media_id(media)
                        bot.logger.info("liked " + display_url + " by " + username)
                        Bots.save_user_info(ig_username, "liked " + display_url + " by " + username)
                        Bots.payment_system()

def main():
	args = get_args()
	mypath = args.db
	output_path = args.output
	img_size = args.img_size
	ad = args.ad

	isPlot = True
	detector = MTCNN()

	onlyfiles_png = []
	onlyfiles_txt = []
	for num in range(0,24):
		if num<9:
			mypath_obj = mypath+'/0'+str(num+1)
		else:
			mypath_obj = mypath+'/'+str(num+1)
		print(mypath_obj)
		onlyfiles_txt_temp = [f for f in listdir(mypath_obj) if isfile(join(mypath_obj, f)) and join(mypath_obj, f).endswith('.txt')]
		onlyfiles_png_temp = [f for f in listdir(mypath_obj) if isfile(join(mypath_obj, f)) and join(mypath_obj, f).endswith('.png')]
	
		onlyfiles_txt_temp.sort()
		onlyfiles_png_temp.sort()

		onlyfiles_txt.append(onlyfiles_txt_temp)

def detect_face(img_path, detector=MTCNN()):
    """
    detect face with MTCNN
    :param img_path:
    :return:
    """
    img = cv2.imread(img_path)
    if detector is None:
        detector = MTCNN()
    mtcnn_result = detector.detect_faces(img)

    return mtcnn_result

ap.add_argument('--gpu', default=0, type=int, help='gpu id')
ap.add_argument('--det', default=0, type=int, help='mtcnn option, 1 means using R+O, 0 means detect from begining')
ap.add_argument('--flip', default=0, type=int, help='whether do lr flip aug')
ap.add_argument('--threshold', default=1.24, type=float, help='ver dist threshold')

args = ap.parse_args()

# Load embeddings and labels
data = pickle.loads(open(args.embeddings, "rb").read())
le = pickle.loads(open(args.le, "rb").read())

embeddings = np.array(data['embeddings'])
labels = le.fit_transform(data['names'])

# Initialize detector
detector = MTCNN()

# Initialize faces embedding model
embedding_model =face_model.FaceModel(args)

# Load the classifier model
model = load_model(args.mymodel)

# Define distance function
def findCosineDistance(vector1, vector2):
    """
    Calculate cosine distance between two vector
    """
    vec1 = vector1.flatten()
    vec2 = vector2.flatten()

    a = np.dot(vec1.T, vec2)