Siamese Neural Network using contrastive learning for face recognition
This code is based on the Siamese NN paper for classification using contrastive learning. After cloning, create 2 folders 'Positive' and 'Negative' and update the corresponding path in the 'SiameseNN.ipynb' file.
The structure looks like this :
Siamese NN
|- Negative
|- Positive
|- SiameseNN.ipynb
Add the image of the class that needs to be learnt in the Positive folder and the images of the classes other than the main class in the Negative folder. Then, run the cells in the notebook and let the model train for 25 epochs. Then evaluate the results.
Note : The images are not added in the repository for privacy purpose