This document presents a person identification system based on iris recognition using deep learning. It uses the ResNet-50 model for feature extraction from iris images. The system first preprocesses iris images to remove noise and reflections. It then segments the iris region using Daugman's operator and normalizes it using his rubber sheet model. The normalized iris images are input to the pre-trained ResNet-50 model to extract features. Finally, multiclass SVM is used to perform person identification based on the extracted features. The system was tested on the VISA iris database and achieved 95% accuracy in identification.