age_model.py

from deepface.basemodels import VGGFace
import os
from pathlib import Path
import gdown
import numpy as np
from keras.models import Model, Sequential
from keras.layers import Convolution2D, Flatten, Activation
from keras.preprocessing import image
import cv2


class Age_Model():
    def __init__(self):
        self.model = self.loadModel()
        self.output_indexes = np.array([i for i in range(0, 101)])
    
    def predict_age(self,face_image):
        image_preprocesing = self.transform_face_array2age_face(face_image)
        age_predictions = self.model.predict(image_preprocesing )[0,:]
        result_age = self.findApparentAge(age_predictions)
        return result_age

    def loadModel(self):
        model = VGGFace.baseModel()
        #--------------------------
        classes = 101
        base_model_output = Sequential()
        base_model_output = Convolution2D(classes, (1, 1), name='predictions')(model.layers[-4].output)
        base_model_output = Flatten()(base_model_output)
        base_model_output = Activation('softmax')(base_model_output)
        #--------------------------
        age_model = Model(inputs=model.input, outputs=base_model_output)
        #--------------------------
        #load weights
        if os.path.isfile('Models/Saved Model/age_model_weights.h5') != True:
            print("age_model_weights.h5 will be downloaded...")
            url = 'https://drive.google.com/uc?id=1YCox_4kJ-BYeXq27uUbasu--yz28zUMV'
            output = 'Models/Saved Model/age_model_weights.h5'
            gdown.download(url, output, quiet=False)
        age_model.load_weights('Models/Saved Model/age_model_weights.h5')
        return age_model
        #--------------------------

    def findApparentAge(self,age_predictions):
        apparent_age = np.sum(age_predictions * self.output_indexes)
        return apparent_age

    def transform_face_array2age_face(self,face_array,grayscale=False,target_size = (224, 224)):
        detected_face = face_array
        if grayscale == True:
            detected_face = cv2.cvtColor(detected_face, cv2.COLOR_BGR2GRAY)
        detected_face = cv2.resize(detected_face, target_size)
        img_pixels = image.img_to_array(detected_face)
        img_pixels = np.expand_dims(img_pixels, axis = 0)
        #normalize input in [0, 1]
        img_pixels /= 255
        return img_pixels