Work

A hybrid deep learning and cryptographic framework using signcryption for robust biometric template protection

Year: 2026

Type: article

Abstract: Protecting biometric data is a crucial concern in the age of digital identity verification. A multimodal template protection system that safeguards iris and fingerprint modalities is presented in this work. Discriminative extraction of features using CNN and then binarized to produce compact binary representations. A robust multimodal feature vector is produced by fusing these binary codes using a Block-XOR method. A signcryption process is used to create protected templates for safe storage and authentication in order to further improve security. Data confidentiality, integrity, and trustworthy authentication are guaranteed by the suggested structure. With low Equal Error Rates of 0.0201%, 0.044%, and 0.496%, respectively, experimental evaluation on three benchmark datasets CMBD, CASIAV1 with FVC2004, and CASIAV3 with FVC2006 shows good performance.

Source: Scientific Reports

DOI: 10.1038/s41598-026-74048-8

Authors: Pritam Kumar Raul, Mulagala Sandhya, Balaprakasa Rao Killi, Dilip Kumar Vallabhadas

Institutions: National Institute of Technology Warangal, SRM University, Andhra Pradesh

Language: English

Topic: Biometric Identification and Security

Subfield: Signal Processing

Field: Computer Science

Domain: Physical Sciences

Open Access status: gold

Estimated APC paid: $2,690

OpenAlex ID: W7220447916