用于法医目的的卷积长短期记忆网络破损指纹识别
计算机视觉与模式识别
2021-01-01 v1 密码学与安全
摘要
指纹识别通常是确立针对罪犯证据的决定性步骤。然而，我们日益发现罪犯以各种方式故意篡改其指纹，使技术人员和自动传感器难以识别其指纹，令调查人员在法医程序中难以确立有力证据。在此意义上，深度学习成为协助识别破损指纹的主要候选方法，尤其是卷积算法。本文聚焦于利用卷积长短期记忆（Convolutional Long Short-Term Memory）网络识别破损指纹。我们给出了模型的架构，并展示其性能：准确率超过95%，精确率99%，召回率接近95%，AUC达99%。
引用
@article{arxiv.2012.15041,
title = {Damaged Fingerprint Recognition by Convolutional Long Short-Term Memory Networks for Forensic Purposes},
author = {Jaouhar Fattahi and Mohamed Mejri},
journal= {arXiv preprint arXiv:2012.15041},
year = {2021}
}
备注
This paper was accepted, on December 5, 2020, for publication and oral presentation at the 2021 IEEE 5th International Conference on Cryptography, Security and Privacy (CSP 2021) to be held in Zhuhai, China during January 8-10, 2021 and hosted by Beijing Normal University (Zhuhai)