Face verification model using Siamese CNNs and One-Shot Learning.
To use this code, modify the inputs section in the Python or Google Colab files (whichever you prefer). The inputs are the following:
- CNN architecture: Choose one of the following: 'LeNet5', 'AlexNet', 'VGG19', or 'GoogleNet'
- Directory: Directory to the training set images.
The directory must contain one folder per person. Each folder must include a minimum of 5 images of that person.
Refer to the 'people' folder in this repository.
- Test directory: Directory to the validation set images.
The directory must include the images that you want to test. It should NOT contain folders.
Refer to the 'test_images' folder in this repository.
- Epochs: Number of training iterations (epochs) in TensorFlow.
