High Fidelity Fingerprint Generation: Quality, Uniqueness, and Privacy
This work addresses the need for synthetic fingerprint datasets for biometric research and security applications, offering a novel method with public availability.
The paper tackles generating realistic high-resolution fingerprints using a progressive growth GAN, resulting in unique and diverse 512x512 pixel fingerprints that preserve privacy by not revealing training data identities.
In this work, we utilize progressive growth-based Generative Adversarial Networks (GANs) to develop the Clarkson Fingerprint Generator (CFG). We demonstrate that the CFG is capable of generating realistic, high fidelity, $512\times512$ pixels, full, plain impression fingerprints. Our results suggest that the fingerprints generated by the CFG are unique, diverse, and resemble the training dataset in terms of minutiae configuration and quality, while not revealing the underlying identities of the training data. We make the pre-trained CFG model and the synthetically generated dataset publicly available at https://github.com/keivanB/Clarkson_Finger_Gen