This Repository is an implementation of Siamese Neural Networks for One-shot Image Recognition by G. Koch et al.
Facial Recognition using Siamese Neural Networks for One-shot Image Recognition. A unique structure is employed to naturally rank similarity between inputs. A convolutional architecture is used to achieve strong results which exceed those of other deep learning models with near state-of-the-art performance on one-shot classification tasks.
The data used for training contains two parts, the positive and the anchor images are collected through webcam while the negative images are taken from Labelled Faces in the Wild Dataset.