Project: facenet-demo (GitHub Link)

facenet-demo-master
- bat_predict.bat
- src
  - download_and_extract.py
  - validate_on_lfw.py
  - freeze_graph.py
  - __pycache__
    - facenet.cpython-36.pyc
    - lfw.cpython-36.pyc
    - detect_face.cpython-36.pyc
  - lfw.py
  - train_tripletloss.py
  - facenet.py
  - models
    - inception_resnet_v1.py
    - __pycache__
      - __init__.cpython-36.pyc
      - inception_resnet_v1.cpython-36.pyc
    - squeezenet.py
    - inception_resnet_v2.py
    - dummy.py
    - __init__.py
  - decode_msceleb_dataset.py
  - generative
    - train_vae.py
    - models
      - dfc_vae_resnet.py
      - vae_base.py
      - dfc_vae_large.py
      - dfc_vae.py
    - calculate_attribute_vectors.py
    - modify_attribute.py
  - align
    - detect_face.py
    - det1.npy
    - align_dataset_mtcnn.py
    - __pycache__
      - facenet.cpython-36.pyc
      - __init__.cpython-36.pyc
      - detect_face.cpython-36.pyc
    - detect_face_images.py
    - facenet.py
    - detect_face_camera.py
  - predict.py
  - export_embeddings.py
  - __init__.py
  - calculate_filtering_metrics.py
  - cluster.py
  - train_softmax.py
  - classifier.py
  - compare.py
- bat_init_env.bat
- bat_compare.bat
- bat_detect_face_camera.bat
- __pycache__
  - __init__.cpython-36.pyc
- tmp
  - align_dlib.py
  - test_invariance_on_lfw.py
  - mtcnn_test.py
  - nn2.py
  - invariance_test.txt
  - seed_test.py
  - mnist_noise_labels.py
  - mtcnn.py
  - nn3.py
  - nn4_small2_v1.py
  - deepdream.py
  - random_test.py
  - vggface16.py
  - vggverydeep19.py
  - mnist_center_loss.py
  - test1.py
  - funnel_dataset.py
  - align_dataset.py
  - download_vgg_face_dataset.py
  - align_dataset.m
  - __init__.py
  - visualize_vggface.py
  - dataset_read_speed.py
  - detect_face_v2.m
  - rename_casia_directories.py
  - visualize.py
  - select_triplets_test.py
  - visualize_vgg_model.py
  - mtcnn_test_pnet_dbg.py
  - network.py
  - cacd2000_split_identities.py
  - test_align.py
  - detect_face_v1.m
  - nn4.py
- test
  - batch_norm_test.py
  - triplet_loss_test.py
  - train_test.py
  - restore_test.py
  - center_loss_test.py
- facenet_api.txt
- util
  - plot_learning_curves.m
- .pydevproject
- .pylintrc
- LICENSE.md
- data
  - learning_rate_retrain_tripletloss.txt
  - images
  - learning_rate_schedule_classifier_vggface2.txt
  - learning_rate_schedule_classifier_casia.txt
  - learning_rate_schedule_classifier_msceleb.txt
  - pairs.txt
- bat_validate_on_lfw.bat
- .travis.yml
- README.md
- bat_classifier_mode_classify.bat
- contributed
  - face.py
  - real_time_face_recognition.py
  - __pycache__
    - facenet.cpython-36.pyc
    - detect_face.cpython-36.pyc
  - predict.py
  - export_embeddings.py
  - batch_represent.py
  - cluster.py
  - clustering.py
- bat_align_dataset_mtcnn.bat
- requirements.txt
- bat_train_tripletloss.bat
- bat_classifier_mode_train.bat
- bat_detect_face_images.bat
- .gitignore
- bat_cluster.bat
- .project
- bat_0description.txt

facenet-demo

数据准备：

①lfw数据集（lfw），bdyun下载：

链接：https://pan.baidu.com/s/1b9gLUEK8NqEiKGzkv0z26Q

提取码：2mhn

存放于 facenet/data

②facenet预训练权重（20170512-110547.pb），bdyun下载：

链接：https://pan.baidu.com/s/1iJjb_45czF5v_cE4uSl24g

提取码：iqhm

存放于 facenet/weight

③facenet预训练pkl文件（classifier.pkl）bdyun下载：

链接：https://pan.baidu.com/s/1TiiRBWxX8a0Kc5ciahmRxg

提取码：oos2

存放于 facenet/weight

运行环境

win10，python3.6.3

运行 bat_init_env.bat 下载依赖库

demo使用

运行 bat_align_dataset_mtcnn.bat 人脸对齐生成 data/lfw_160

运行 bat_detect_face_images.bat 静态图人脸检测、人脸关键点检测

运行 bat_detect_face_camera.bat 摄像头人脸检测、人脸关键点检测

运行 bat_compare.bat 人脸对比

运行 bat_cluster.bat 人脸聚类

运行 bat_predict.bat 在人脸库查找某个人脸

运行 bat_train_tripletloss.bat 训练网络