Facenet for face recognition using pytorch

• Pytorch implementation of the paper: "FaceNet: A Unified Embedding for Face Recognition and Clustering".
• Training of network is done using triplet loss.
• I will update codes and git repository for a better accuracy after my doctoral defence on Nov. 20th 2018.

How to train/validate model

• Download vggface2 (for training) and lfw (for validation) datasets.

  • https://www.robots.ox.ac.uk/%7Evgg/data/vgg_face2/
  • http://vis-www.cs.umass.edu/lfw/lfw.tgz

• Align face image files by following David Sandberg's instruction (part of "Face alignment").

  • https://github.com/davidsandberg/facenet/wiki/Classifier-training-of-inception-resnet-v1

• Write list file of face images by running "datasets/write_csv_for_making_dataset.ipynb"

  • This is aready in the directory of datasets so that you don't need to do again if you are urgent.
  • To run this one need to modify paths in accordance with location of image dataset.

• Train

  • Again, one need to modify paths in accordance with location of image dataset.
  • Also feel free to change some parameters.

Results

• Accuracy on VGGFace2 and LFW datasets

• Triplet loss on VGGFace2 and LFW datasets

• ROC curve on LFW datasets for validation

References

• https://github.com/liorshk/facenet_pytorch
• https://github.com/davidsandberg/facenet
• https://arxiv.org/abs/1503.03832