One-Shot-Learning-and-Siamese-Network
One-shot learning allows deep learning algorithms to measure the similarity and difference between two images.
1?Take an input and extract its embedding (mapping to a vector of continuous numbers) by passing it through a neural network. 2>Repeat step 1 with a different input. 3>Compare the two embeddings to check whether there is a similarity between the two data points. These two embeddings act as a latent feature representation of the data. In our case, images with the same person should have similar embeddings.
note: download the facenet.h5 before using this project.
#Reference https://github.com/susantabiswas/FaceRecog