# coding:utf-8
import dlib
import numpy as np
from copy import deepcopy
import cv2
import os
from LiveDetection import detect_live
import torch

# http://dlib.net/files/dlib_face_recognition_resnet_model_v1.dat
# http://dlib.net/files/shape_predictor_68_face_landmarks.dat.bz2


class FaceRecognition(object):
    def __init__(self, input_img,face_vector_model):
        super(FaceRecognition, self).__init__()
        self.input_img = input_img
        self.detector = dlib.get_frontal_face_detector()
        self.img_size = 150
        self.predictor = dlib.shape_predictor(r'./shape_predictor_68_face_landmarks.dat')
        self.recognition = dlib.face_recognition_model_v1('dlib_face_recognition_resnet_model_v1.dat')
        self.live_threshold = 0.8
        self.face_vector_model=face_vector_model


    def point_draw(self, img, sp, title, save):
        img = cv2.cvtColor(img, cv2.COLOR_RGB2BGR)
        for i in range(68):
            cv2.putText(img, str(i), (sp.part(i).x, sp.part(i).y), cv2.FONT_HERSHEY_DUPLEX, 0.3, (0, 0, 255), 1,
                        cv2.LINE_AA)
            # cv2.drawKeypoints(img, (sp.part(i).x, sp.part(i).y),img, [0, 0, 255])
        if save:
            # filename = title+str(np.random.randint(100))+'.jpg'
            filename = title + '.jpg'
            cv2.imwrite(filename, img)
        # os.system("open %s"%(filename))
        # cv2.imshow(title, img)
        # cv2.waitKey(0)
        # cv2.destroyWindow(title)

    def show_origin(self, img):
        cv2.imshow('origin', img)
        cv2.waitKey(0)
        cv2.destroyWindow('origin')

    def getfacefeature(self, image, check_live=False):
        # import pdb
        # pdb.set_trace()
        # image = dlib.load_rgb_image(img)
        ## 人脸对齐、切图
        # 人脸检测
        dets = self.detector(image, 1)
        if len(dets) == 1:
            # faces = dlib.full_object_detections()
            # 关键点提取
            shape = self.predictor(image, dets[0])
            print("Computing descriptor on aligned image ..")
            # 人脸对齐 face alignment
            images = dlib.get_face_chip(image, shape, size=self.img_size)

            # self.point_draw(image, shape, 'before_' + img, save=True)
            shapeimage = np.array(images).astype(np.uint8)
            dets = self.detector(shapeimage, 1)
            if len(dets) == 1:
                point68 = self.predictor(shapeimage, dets[0])
                # self.point_draw(shapeimage, point68, 'after_' + img, save=True)
                #     Live detection
                if check_live:
                    resized_img = cv2.resize(image, (64, 64))
                    prob = detect_live(resized_img)
                    if prob < self.live_threshold:
                        return None
                    # 计算对齐后人脸的128维特征向量
                # face_descriptor_from_prealigned_image = self.recognition.compute_face_descriptor(images)
                model = torch.load(self.face_vector_model)
                face_vector = model(images)
                return face_vector
            else:
                return None
        else:
            return None

    def compare(self):
        # import pdb
        # pdb.set_trace()
        vec1 = self.getfacefeature(self.input_img, check_live=True)
        if vec1 is None:
            print('Not Real Person')
            return False
        else:
            print("Have already detect live person")
            vec1 = np.array(vec1)
            vec2 = np.array(self.getfacefeature(self.src_img))

            # import pdb
            # pdb.set_trace()
            dest = np.sqrt(np.sum((vec1 - vec2) * (vec1 - vec2)))
            print('distance between people:{:.3f}'. \
                  format(dest))
            return True

    def getFaceEmbeddingVectors(self):
        model = torch.load('\model.pkl')


def compareFromCamera(src_face_vector):
    # 导入OpenCV
    import cv2
    # 创建一个VideoCapture对象，它的参数可以是设备索引或视频文件的名称（下面会讲到）。设备索引只是指定哪台摄像机的号码。0代表第一台摄像机、1代表第二台摄像机。之后，可以逐帧捕捉视频。最后释放捕获。
    cap = cv2.VideoCapture(0)
    while True:
        # 读取帧
        ret, frame = cap.read()

        # 将视频灰度化
        # gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
        # 将视频灰度化显示
        cv2.imshow('frame', frame)
        detection_recognition = FaceRecognition( frame)
        if detection_recognition.compare():
            break
        # 按‘q'退出
        if cv2.waitKey(1) and 0xFF == ord('q'):
            break
    # 释放资源并关闭窗口
    cap.release()
    cv2.destroyAllWindows()


if __name__ == '__main__':
    my_face_vector= [ 1.6443e-01,  2.1826e-01,  3.4180e-01, -3.3691e-01,  7.7576e-02,
         -4.1290e-02,  2.1362e-01, -2.6758e-01,  2.7637e-01, -5.0391e-01,
         -6.9092e-02, -2.3303e-01,  5.2826e-02, -2.1216e-01,  1.0907e-01,
          4.9170e-01,  2.3499e-01,  1.2769e-01, -3.9697e-01,  2.2595e-01,
         -3.5327e-01,  8.9893e-01,  1.1609e-01,  2.1021e-01,  2.9370e-01,
          3.1201e-01,  4.2285e-01,  5.8398e-01, -7.8796e-02, -2.0471e-01,
         -1.3208e-01,  4.5142e-01, -1.2671e-01,  1.7200e-01, -1.2720e-01,
         -1.6406e-01, -2.0276e-01, -6.0645e-01, -4.4849e-01, -2.3743e-01,
         -6.4258e-01, -1.0908e+00,  1.2830e-01,  4.0674e-01,  2.7734e-01,
         -6.0272e-02, -3.4546e-01,  4.4141e-01, -9.8343e-03,  5.5322e-01,
          4.1162e-01,  3.0420e-01,  4.0552e-01,  1.0236e-01, -2.7368e-01,
          3.3887e-01, -1.8982e-01, -2.4695e-01, -1.4148e-01,  3.3350e-01,
          1.4343e-01, -4.2114e-01, -1.6931e-01, -3.7427e-01,  3.5938e-01,
          1.9556e-01, -1.6394e-01,  2.7466e-01,  1.1414e-01, -2.9419e-01,
         -6.4844e-01, -3.4692e-01, -6.7627e-01,  2.7954e-01, -6.9763e-02,
          2.8540e-01, -1.0144e-01,  2.5952e-01, -2.7612e-01, -4.5166e-01,
         -3.7744e-01, -6.3818e-01, -1.2978e-02,  2.9688e-01, -1.6089e-01,
         -1.1652e-01, -1.5784e-01, -2.9492e-01,  2.1497e-01,  2.6074e-01,
         -6.4893e-01, -4.8730e-01, -4.6680e-01, -1.8341e-02,  3.6182e-01,
         -2.8320e-02, -1.5211e-03, -3.6426e-01, -2.4500e-01, -2.9956e-01,
          4.6997e-01,  4.2773e-01,  5.3101e-02, -5.3027e-01, -3.3081e-01,
         -2.0618e-01,  5.2094e-02,  5.2881e-01, -5.1025e-01,  6.6797e-01,
         -1.8103e-01, -2.6108e-02, -2.8247e-01,  6.0156e-01, -2.7979e-01,
         -1.0181e-01, -6.2683e-02, -2.8003e-01,  3.0457e-02,  3.4399e-01,
         -6.5820e-01,  2.8839e-02,  8.1848e-02,  4.7388e-01,  2.0837e-01,
          3.5522e-01, -2.0288e-01,  3.3691e-01,  1.5955e-01, -4.2944e-01,
         -3.1982e-01, -2.5952e-01,  8.2092e-02,  2.4976e-01,  4.2578e-01,
         -2.4933e-02, -8.4045e-02,  4.5044e-01, -9.8938e-02, -6.0107e-01,
         -1.3440e-01, -1.4172e-01,  2.6685e-01, -1.5088e-01,  1.0553e-01,
          4.1870e-01, -2.6611e-01,  4.8633e-01, -1.6205e-02, -9.9670e-02,
          1.3037e-01, -4.7437e-01,  1.6711e-01, -6.6528e-02,  1.5671e-02,
          7.1533e-01,  9.6313e-02, -3.3594e-01, -3.7695e-01, -3.3154e-01,
         -8.6975e-02, -2.0813e-01,  3.2715e-01,  5.1758e-02, -3.9282e-01,
          3.0029e-01, -3.2837e-01,  2.9144e-02, -1.5511e-02, -1.5833e-01,
         -3.8940e-01,  8.8318e-02, -2.0715e-01, -2.8662e-01, -6.4453e-02,
          3.2153e-01, -3.9380e-01, -4.2212e-01, -7.0129e-02,  1.9324e-01,
         -2.1716e-01, -2.3584e-01,  1.8567e-01, -2.9785e-01,  9.5032e-02,
          1.9806e-02,  2.6099e-01, -2.3389e-01,  2.6025e-01, -5.0098e-01,
          1.2634e-01,  2.0911e-01,  5.8691e-01,  1.9394e-02,  1.1151e-01,
          1.3477e-01,  1.8359e-01, -4.9536e-01,  1.2646e-01, -1.1493e-01,
         -8.2886e-02,  5.6091e-02,  6.0791e-01,  3.2568e-01,  7.1289e-01,
          9.9792e-02, -1.5662e-01, -3.7280e-01,  2.2888e-01,  4.7949e-01,
          1.6663e-01, -1.9702e-01, -3.8128e-03, -2.2754e-01, -5.3662e-01,
         -1.6211e-01, -5.9521e-01, -2.4280e-01,  2.6660e-01, -1.3647e-01,
         -4.8071e-01, -2.1899e-01, -5.1367e-01, -6.7383e-01,  3.2104e-01,
          1.9531e-01,  1.8204e-02, -1.4343e-01,  3.1299e-01,  1.2903e-01,
          7.2876e-02, -1.8187e-03, -3.5547e-01,  6.9702e-02,  2.0300e-01,
         -2.2095e-02, -2.0630e-01, -1.9507e-01, -1.1871e-01,  2.2253e-01,
         -1.1603e-01,  1.5979e-01,  4.3530e-01,  8.6365e-02, -3.6792e-01,
         -6.4307e-01,  1.9495e-01, -1.4001e-01,  9.4528e-03, -1.4197e-01,
         -1.1969e-01,  6.3330e-01, -2.4512e-01,  4.5776e-01,  2.7319e-01,
          4.3286e-01, -5.4199e-02,  4.0796e-01,  1.5344e-01,  4.2676e-01,
          4.5319e-02,  5.8740e-01, -2.8027e-01, -2.1204e-01, -1.2756e-01,
         -3.0322e-01, -6.6260e-01,  1.2939e-01, -4.2139e-01,  3.9722e-01,
         -4.2017e-01,  3.2910e-01, -3.2153e-01,  5.7227e-01, -4.4678e-02,
         -1.3550e-01, -3.8477e-01, -2.4185e-02, -5.1178e-02, -2.9614e-01,
         -5.7861e-01, -5.4346e-01,  3.3789e-01, -1.2866e-01, -1.1542e-01,
         -3.0322e-01, -4.2358e-01, -6.2354e-01,  5.7910e-01, -2.6901e-02,
         -3.2617e-01, -4.6570e-02,  5.2051e-01, -3.3789e-01,  3.2324e-01,
         -3.7659e-02,  2.8369e-01,  1.1945e-01,  2.7618e-02,  1.5906e-01,
         -3.3984e-01, -3.3594e-01,  1.4404e-01, -3.8501e-01, -8.4167e-02,
         -2.1350e-01, -4.6216e-01,  1.3037e-01,  2.1594e-01, -5.7471e-01,
         -5.0146e-01,  6.4990e-01, -4.7998e-01,  5.2246e-01, -4.3359e-01,
         -9.8999e-02, -2.9590e-01,  1.8298e-01,  3.8849e-02,  3.4155e-01,
         -3.9844e-01, -3.7384e-04, -1.6815e-02, -2.0776e-01,  2.4475e-01,
         -4.6558e-01,  2.9858e-01, -2.8174e-01, -2.0081e-01, -3.1934e-01,
          1.1890e-01,  1.6739e-02,  1.1981e-01, -3.9941e-01,  1.3092e-02,
          2.1313e-01, -3.1226e-01,  5.3101e-02,  2.3206e-01,  8.7158e-01,
         -4.6899e-01,  2.5269e-01,  1.3611e-01, -6.3232e-01, -1.8250e-01,
          1.1499e-01,  1.9495e-01,  3.7988e-01,  3.0640e-01,  7.0129e-02,
         -4.3274e-02,  5.3528e-02, -5.3613e-01,  1.9128e-01, -2.9419e-01,
          3.3154e-01,  2.1545e-01, -2.3499e-01,  2.1118e-01,  2.4146e-01,
         -1.3452e-01, -1.6736e-01,  3.5547e-01, -4.4287e-01,  1.6205e-02,
         -2.5195e-01,  3.8013e-01, -3.5742e-01, -4.5142e-01,  2.2949e-01,
          5.7324e-01,  3.5449e-01,  3.9990e-01, -1.2793e-01,  1.6016e-01,
          2.5000e-01,  3.9233e-01, -3.7915e-01,  7.9407e-02,  4.6924e-01,
         -2.8442e-01, -3.0396e-01, -4.4971e-01, -3.5889e-01,  1.6541e-01,
          4.1064e-01,  3.2593e-01,  4.0283e-01,  2.6172e-01,  1.3928e-01,
          1.1432e-01, -2.7710e-01,  6.4209e-01,  3.8696e-01, -4.3373e-03,
         -1.3269e-01, -2.9846e-02, -1.6937e-02,  1.8054e-01, -2.9834e-01,
         -9.9426e-02,  4.9390e-01, -1.3269e-01, -2.7222e-01, -7.3364e-02,
          2.9761e-01,  2.9248e-01,  2.5439e-01, -2.5708e-01,  2.2571e-01,
         -2.1069e-01, -3.3154e-01, -1.3342e-01,  2.6538e-01, -1.1902e-01,
          1.8945e-01, -3.8672e-01, -3.6816e-01,  3.4668e-01, -3.5547e-01,
          6.1670e-01, -2.2668e-01,  2.2974e-01,  8.4595e-02,  5.2734e-01,
          4.8120e-01,  3.0835e-01,  2.1716e-01,  3.8574e-02,  2.8052e-01,
         -1.1481e-01,  2.5537e-01,  2.6904e-01, -3.9697e-01, -7.3120e-02,
         -1.4709e-01,  2.2400e-01, -1.2535e-02,  6.4307e-01, -4.7168e-01,
         -3.1885e-01, -5.2441e-01, -4.2725e-01, -2.3239e-02,  3.5669e-01,
          3.7744e-01, -7.8760e-01, -7.4219e-02,  5.7080e-01, -2.1594e-01,
          1.7590e-01,  6.9519e-02, -2.2815e-01, -3.2422e-01,  2.8101e-01,
         -5.1221e-01,  3.0591e-01, -5.2393e-01, -4.8535e-01,  3.8721e-01,
         -3.0014e-02, -2.2263e-02, -2.4780e-01,  3.9307e-01, -1.3603e-02,
          9.7107e-02, -5.2344e-01, -2.9175e-01, -1.4343e-01, -4.0088e-01,
         -2.0782e-02, -2.1680e-01,  3.6499e-01, -1.4502e-01,  5.2588e-01,
          4.3213e-01,  3.8574e-01,  1.0300e-02,  7.6599e-02,  3.8818e-02,
         -7.9651e-02, -1.7603e-01, -3.3142e-02,  5.3528e-02, -7.9041e-02,
         -5.8594e-01, -1.3208e-01, -3.5950e-02, -8.2886e-02,  6.1035e-02,
         -6.8945e-01,  6.0944e-02, -2.6465e-01, -4.9951e-01,  3.3228e-01,
         -3.6865e-01, -1.2488e-01,  9.4452e-03,  5.5859e-01,  1.7944e-01,
         -3.1104e-01, -2.6709e-01, -1.0553e-01, -5.8740e-01,  1.8280e-02,
         -3.8788e-02,  4.5850e-01,  2.1350e-01,  1.2854e-01, -5.7831e-02,
         -4.5776e-01, -2.0032e-01,  1.2427e-01,  4.7656e-01, -2.1045e-01,
         -2.9850e-04, -7.6416e-02, -4.8584e-01, -3.7048e-02, -6.8359e-02,
          6.9287e-01,  7.0435e-02, -8.0933e-02, -4.5825e-01, -4.5746e-02,
          1.2769e-01, -4.6600e-02,  6.5979e-02, -1.1670e-01, -3.0225e-01,
          1.4551e-01,  8.0627e-02, -2.4072e-01, -4.9194e-01, -6.4746e-01,
          1.0162e-01, -3.4448e-01,  1.1163e-01, -3.1396e-01, -1.6199e-01,
          3.9136e-01,  2.1988e-02,  1.8933e-01, -9.1370e-02,  3.9722e-01,
          2.2119e-01,  4.6851e-01, -5.9753e-02, -2.7783e-01, -3.7036e-01,
          2.7075e-01,  4.4507e-01, -1.1560e-01,  2.9816e-02, -4.9097e-01,
         -4.5703e-01,  1.8631e-02, -2.1667e-01, -3.0411e-02,  1.0950e-01,
         -6.3086e-01, -2.1106e-01, -8.6731e-02, -2.2449e-01, -2.3242e-01,
          1.8286e-01,  2.5781e-01,  2.2571e-01, -3.7012e-01, -7.6050e-02,
          9.1003e-02, -1.8884e-01,  2.6099e-01, -9.3002e-03,  1.4183e-02,
          3.3716e-01,  4.7925e-01,  8.3862e-02, -1.3684e-01,  4.0112e-01,
         -3.7500e-01, -4.6851e-01,  2.2485e-01,  2.9761e-01, -7.5073e-02,
         -2.6245e-01, -3.4332e-02,  3.3301e-01,  5.4932e-01, -2.9810e-01,
          1.1298e-01,  4.4629e-01, -1.1462e-01,  2.6001e-01, -2.3804e-01,
         -1.8005e-01, -4.9707e-01,  2.4817e-01, -2.6733e-01,  1.8738e-01,
          2.7313e-03,  4.5728e-01, -4.6997e-01,  1.3748e-02, -2.0898e-01,
          1.8127e-01,  1.5930e-01,  5.2441e-01, -2.3880e-02, -3.8525e-01,
          5.6104e-01,  2.3352e-01, -2.4854e-01, -4.4653e-01,  2.1619e-01,
         -5.3040e-02, -2.9510e-02, -5.3809e-01, -2.0032e-01, -3.6652e-02,
          2.4524e-01, -2.6440e-01, -2.1570e-01,  4.6655e-01,  2.7405e-02,
         -3.5352e-01,  6.8726e-02, -1.9287e-01, -3.5059e-01, -4.5557e-01,
         -5.0879e-01,  2.7246e-01,  3.5010e-01,  6.8481e-02, -4.2871e-01,
          4.2334e-01, -2.4231e-01,  2.6636e-01,  1.3770e-01,  3.4766e-01,
          2.5122e-01,  2.0581e-01,  2.9150e-01,  1.2262e-01,  2.0679e-01,
         -4.5142e-01, -5.5273e-01,  4.9512e-01,  3.9966e-01,  2.1838e-01,
         -1.9971e-01,  2.4133e-01,  1.6553e-01,  5.3516e-01,  1.9714e-01,
          5.8838e-01,  1.1621e-01, -1.2177e-01,  1.0260e-01,  2.2461e-01,
          2.7222e-01,  2.2095e-01,  3.6060e-01,  2.6050e-01,  6.0352e-01,
          1.5112e-01, -3.6865e-01,  1.0291e-01,  4.5996e-01,  1.3293e-01,
         -5.0926e-03,  5.3955e-01, -1.7761e-01,  3.9575e-01, -3.9703e-02,
          4.0576e-01, -2.6074e-01,  3.1714e-01, -4.5728e-01, -5.2277e-02,
         -3.1812e-01,  3.8605e-02, -7.4121e-01,  2.7490e-01,  1.9824e-01,
          3.2520e-01, -1.9080e-01,  3.0444e-01,  2.9883e-01, -2.1582e-01,
         -4.7852e-01,  6.4636e-02, -2.5757e-01,  3.4814e-01, -3.0054e-01,
         -1.7395e-01, -3.4473e-01, -2.8271e-01,  7.2266e-02, -2.5244e-01,
          3.1738e-01, -4.5264e-01,  1.9360e-01,  1.0602e-01, -5.0488e-01,
         -6.1377e-01,  1.3000e-01,  3.9990e-01,  1.2219e-01, -1.9775e-01,
         -2.3438e-02,  4.2529e-01, -1.5308e-01,  1.9287e-01,  5.8533e-02,
          3.1079e-01,  8.5388e-02, -1.9336e-01, -2.4133e-01,  5.4541e-01,
         -9.9426e-02, -5.3955e-01, -3.0041e-03,  1.0455e-01,  4.4482e-01,
          3.9478e-01, -2.9004e-01,  9.6985e-02,  4.7778e-01,  3.8086e-01,
         -8.7036e-02,  2.3694e-01,  5.8545e-01, -3.0396e-01,  7.4414e-01,
          2.2034e-01,  6.0498e-01,  4.5239e-01, -5.2490e-02, -6.3965e-01,
         -4.1650e-01,  2.5610e-01,  3.4180e-01, -1.8140e-01, -2.9556e-02,
          9.0881e-02, -2.2009e-01, -5.8350e-02,  2.0349e-01,  6.0303e-01,
         -2.2852e-01,  4.6851e-01, -3.3234e-02, -6.2891e-01,  5.7666e-01,
         -3.5815e-01,  5.0439e-01,  5.1514e-01, -3.7207e-01,  7.5317e-02,
         -6.2622e-02,  6.5332e-01, -3.8135e-01, -3.2690e-01, -8.8318e-02,
         -3.1815e-03,  2.7344e-01, -5.2460e-02,  6.2408e-02,  3.0334e-02,
          2.5436e-02, -1.6113e-01, -3.9825e-02,  3.2495e-01,  1.0712e-01,
         -3.2153e-01,  5.1660e-01, -1.2231e-01, -1.7993e-01, -1.6406e-01,
          2.6367e-01,  2.5854e-01,  2.2974e-01, -1.7053e-01, -2.8778e-02,
          4.8413e-01, -6.4758e-02,  4.5776e-02,  5.0244e-01, -5.0439e-01,
          9.5581e-02,  3.2251e-01,  1.0461e-01, -2.7954e-01, -5.1709e-01,
          3.1519e-01,  5.2832e-01,  3.4714e-03, -1.1115e-01,  1.1407e-01,
         -3.4424e-01,  1.2573e-01, -1.0980e-01, -3.7134e-01,  3.3105e-01,
          2.0129e-01,  1.0541e-01,  5.4199e-01, -1.1176e-01,  5.6946e-02,
         -2.6050e-01,  2.2852e-01,  4.0112e-01,  1.6321e-01, -3.5840e-01,
          2.1252e-01, -1.9470e-01,  2.2314e-01,  7.0410e-01, -3.5425e-01,
          2.0862e-01,  2.1826e-01,  1.2000e-01,  2.5854e-01, -1.9519e-01,
          9.6069e-02, -4.3018e-01, -3.9282e-01,  6.6833e-02, -9.1431e-02,
         -4.7681e-01, -3.1641e-01, -3.8916e-01, -4.0039e-01,  1.9604e-01,
         -4.7485e-02,  2.5928e-01,  4.4360e-01, -4.5020e-01,  4.4873e-01,
         -3.4943e-02, -5.0146e-01,  1.0388e-01, -2.8979e-01, -5.0146e-01,
         -2.6660e-01, -4.7388e-01, -4.2206e-02,  3.9233e-01, -1.6162e-01,
          7.4609e-01, -9.6375e-02, -2.9053e-01, -3.2031e-01,  3.7670e-03,
         -2.1021e-01, -1.1993e-02, -2.5830e-01,  2.4817e-01, -9.5886e-02,
          3.9978e-02, -2.0398e-01, -1.2610e-01, -5.2441e-01,  1.1359e-01,
          3.7866e-01,  4.7656e-01,  2.3877e-01, -2.9053e-01, -1.7041e-01,
          3.4912e-02, -2.5098e-01,  1.3037e-01, -3.4692e-01,  3.5864e-01,
          2.1130e-01, -3.9648e-01, -3.6743e-01,  9.5764e-02, -1.6675e-01,
          3.8135e-01, -4.5532e-02,  3.9886e-02, -1.0687e-01, -2.8223e-01,
         -2.4963e-01,  3.8818e-01,  3.2153e-01,  1.2402e-01, -3.0859e-01,
          4.5630e-01, -1.9043e-01,  3.8818e-01, -5.6396e-01,  3.9258e-01,
          4.2383e-01, -1.7542e-01,  3.7549e-01,  2.6758e-01, -1.8237e-01,
         -2.4268e-01, -3.1641e-01, -4.0039e-01,  1.3684e-01, -7.0618e-02,
          3.7817e-01, -4.6240e-01,  2.3938e-01, -1.4038e-01, -9.6741e-02,
         -1.7786e-01, -1.2317e-01,  2.2131e-01, -1.1407e-01, -1.2585e-01,
          5.9540e-02, -6.2012e-02, -1.2152e-01, -2.3047e-01,  6.5422e-03,
          2.1021e-01, -1.7810e-01,  3.4497e-01,  8.3435e-02, -2.0447e-01,
          2.0435e-01,  2.8027e-01,  2.3401e-01,  2.9907e-01, -2.1667e-01,
          1.7542e-01,  4.0398e-03,  3.3496e-01,  3.6469e-02, -2.4463e-01,
         -4.5117e-01, -2.5903e-01, -5.1727e-02,  2.5708e-01,  1.7175e-01,
         -4.1992e-01, -1.5222e-01,  3.5840e-01,  7.9834e-02,  5.0537e-01,
          1.0992e-01,  1.2720e-01,  1.9250e-01,  4.3237e-01,  7.5562e-02,
          3.8135e-01,  1.4050e-01, -3.0103e-01,  3.0908e-01,  6.0693e-01,
          8.8882e-03,  2.3779e-01,  3.7744e-01, -4.3604e-01,  2.9297e-01,
          8.6304e-02, -2.4829e-01,  5.8740e-01,  5.2832e-01, -9.9792e-02,
         -2.3145e-01,  3.5889e-01, -7.2937e-02,  4.5142e-01,  4.5825e-01,
         -2.4255e-01,  8.2855e-03, -3.1812e-01, -4.4556e-01,  1.7639e-01,
         -1.6638e-01,  1.3147e-01,  1.4038e-01, -3.4821e-02,  2.2522e-01,
          1.5674e-01,  3.4729e-02, -5.4443e-01,  9.2590e-02,  4.0845e-01,
         -2.2003e-02, -3.7183e-01, -1.3147e-01, -4.9365e-01,  3.6182e-01,
         -1.6394e-01, -2.4805e-01,  6.9519e-02,  2.8540e-01, -7.3051e-03,
          4.8462e-01,  5.3027e-01,  2.1448e-01,  1.0522e-01, -5.2295e-01]
    compareFromCamera(my_face_vector,'models/cbam18.pkl')