Iris Verification with Convolutional Neural Network

We propose a novel convolutional neural network to verify a match between two images of the human iris. The network is trained end-to-end and validated on three publicly available datasets yielding state-of-the-art results against four baseline methods. The network performs better by a10%marginto the state-of-the-art method on the CASIA.v4 dataset. In the network, we use a novel layer whose output is interpreted as a normalized response in the complex plane. We show that the layer improves the performance of the model up to15%on previously-unseen data.