Iris-recognition

Iris Recognition using deep learning methods on jpeg compressed images.

PythonDeep learning

High-level design

A biometric recognition pipeline: extract the iris region, then classify it with a deep learning model — trained specifically on JPEG-compressed input, a harder and more realistic case than raw sensor images.

iris segmentation

isolates the iris region from the eye image

↓

deep learning classifier

matches/recognizes the iris — 98%+ accuracy