Things to include:

  • liveness detection
  • false/ spoof face detection and ignorence
  • face tracking
  • self improving and making its own datasets
  • preserve detailed log file

Flow of program

  • list files in data directory -done
  • extract frontal face (harcascade -done | )
  • resize all images to same size 150x150x1
  • augment positive dataset
  • generate test and train object (test_images, test_labels, train_images, train_labels)