FingerNet: Pushing The Limits of Fingerprint Recognition Using Convolutional Neural Network

TL;DR

This paper introduces FingerNet, an end-to-end CNN framework for fingerprint recognition that jointly learns features and recognition, achieving high accuracy and providing visual explanations of important fingerprint regions.

Contribution

The paper presents a novel deep learning framework that improves fingerprint recognition accuracy and offers visualization of key features, advancing biometric recognition technology.

Findings

Achieved high recognition accuracy on a benchmark dataset.

Outperformed previous fingerprint recognition approaches.

Provided visualizations highlighting important fingerprint regions.

Abstract

Fingerprint recognition has been utilized for cellphone authentication, airport security and beyond. Many different features and algorithms have been proposed to improve fingerprint recognition. In this paper, we propose an end-to-end deep learning framework for fingerprint recognition using convolutional neural networks (CNNs) which can jointly learn the feature representation and perform recognition. We train our model on a large-scale fingerprint recognition dataset, and improve over previous approaches in terms of accuracy. Our proposed model is able to achieve a very high recognition accuracy on a well-known fingerprint dataset. We believe this framework can be widely used for biometrics recognition tasks, making more scalable and accurate systems possible. We have also used a visualization technique to highlight the important areas in an input fingerprint image, that mostly impact the recognition results.