人脸录入录入不同的人脸

我想用dlib录入两张不同的人脸，生成两个不同的csv文件，并识别出来，应该怎么做？

可以参考一下 觉得有帮助可以采纳一下
导入dlib和numpy库:
python

import dlib
import numpy as np

2. 创建dlib的人脸检测器和预测器:

detector = dlib.get_frontal_face_detector()
predictor = dlib.shape_predictor('shape_predictor_68_face_landmarks.dat')

3. 创建两个CSV文件来存储人脸特征:

csv_file1 = open('face_data1.csv', 'w') 
csv_file2 = open('face_data2.csv', 'w')

4. 读取两张人脸图片,检测人脸并提取特征:

img1 = dlib.load_rgb_image('face1.jpg')
dets1 = detector(img1, 1)
for k, d in enumerate(dets1):
    shape1 = predictor(img1, d)
    face_descriptor1 = face_recognizer.compute_face_descriptor(img1, shape1)
np.savetxt(csv_file1, face_descriptor1) 

img2 = dlib.load_rgb_image('face2.jpg')
dets2 = detector(img2, 1) 
for k, d in enumerate(dets2):
    shape2 = predictor(img2, d)
    face_descriptor2 = face_recognizer.compute_face_descriptor(img2, shape2)
np.savetxt(csv_file2, face_descriptor2)

5. 读取测试图片,提取人脸特征并与两张人脸进行比较,识别出最相似的人脸:

   test_img = dlib.load_rgb_image('test_face.jpg')
   dets = detector(test_img, 1)
   for k, d in enumerate(dets):
    shape = predictor(test_img, d)
    face_descriptor = face_recognizer.compute_face_descriptor(test_img, shape)
   face_data1 = np.loadtxt('face_data1.csv')    
   face_data2 = np.loadtxt('face_data2.csv')
   dist1 = np.linalg.norm(face_data1 - face_descriptor, axis=1) 
   dist2 = np.linalg.norm(face_data2 - face_descriptor, axis=1) 
   if min(dist1) < min(dist2):
    print('Face 1') 
   else: 
    print('Face 2')