LEADER: Lightweight End-to-End Attention-Gated Dual Autoencoder for Robust Minutiae Extraction

TL;DR

LEADER is a lightweight, end-to-end neural network for fingerprint minutiae extraction that achieves state-of-the-art accuracy, robust cross-domain performance, and high efficiency with only 0.9M parameters.

Contribution

This paper introduces LEADER, a novel attention-gated dual autoencoder architecture that performs complete minutiae extraction directly from raw fingerprint images in an end-to-end manner.

Findings

Achieves 34% higher F1-score on NIST SD27 compared to specialized extractors.

Secures first place in 47% of challenging latent fingerprint samples.

Runs inference in 15ms on GPU and 322ms on CPU, outperforming commercial software.

Abstract

Minutiae extraction, a fundamental stage in fingerprint recognition, is increasingly shifting toward deep learning. However, truly end-to-end methods that eliminate separate preprocessing and postprocessing steps remain scarce. This paper introduces LEADER (Lightweight End-to-end Attention-gated Dual autoencodER), a neural network that maps raw fingerprint images to minutiae descriptors, including location, direction, and type. The proposed architecture integrates non-maximum suppression and angular decoding to enable complete end-to-end inference using only 0.9M parameters. It employs a novel "Castle-Moat-Rampart" ground-truth encoding and a dual-autoencoder structure, interconnected through an attention-gating mechanism. Experimental evaluations demonstrate state-of-the-art accuracy on plain fingerprints and robust cross-domain generalization to latent impressions. Specifically, LEADER attains a 34% higher F1-score on the NIST SD27 dataset compared to specialized latent minutiae extractors. Sample-level analysis on this challenging benchmark reveals an average rank of 2.07 among all compared methods, with LEADER securing the first-place position in 47% of the samples-more than doubling the frequency of the second-best extractor. The internal representations learned by the model align with established fingerprint domain features, such as segmentation masks, orientation fields, frequency maps, and skeletons. Inference requires 15ms on GPU and 322ms on CPU, outperforming leading commercial software in computational efficiency. The source code and pre-trained weights are publicly released to facilitate reproducibility.