A Width-Matched Comparison of Hybrid Quantum-Classical Self-Supervised Learning for Fingerprint Recognition
Abstract
Fingerprint recognition is a widely deployed biometric, but supervised training requires large labeled enrollment sets. Self-supervised learning (SSL) removes this requirement, and hybrid quantum-classical models have been proposed to enrich the learned representations. Prior quantum SSL studies consider a single contrastive objective, so it is unclear whether reported benefits depend on the objective or can be attributed to the quantum circuit. We insert the QuFeX quantum feature-extraction module into three SSL frameworks, the contrastive SimCLR and MoCo v2 and the non-contrastive BYOL, and compare each hybrid with its classical counterpart at matched representation width (8 features, equal to 8 qubits) on the SOCOFing fingerprint dataset, with a CIFAR-10 control, using k-nearest-neighbor identification on encoder features. In single-run experiments the hybrid scores clearly higher for both contrastive objectives, whereas for BYOL a multi-seed analysis shows no reliable difference, suggesting that any benefit depends on the SSL objective. A hardware-efficient circuit (QNet) does not show the same gain. We examine whether the gains can be attributed to the quantum circuit, considering circuit architecture, trainable parameter count, nonlinearity, and the classical simulability of 8-qubit circuits.
Related
- Spectral fingerprint recognition of mobile signals via Rydberg atomic sensors
- Comment on "Supervised Learning with Quantum Hamiltonian Embeddings" (arXiv:2610.26833)
- Classical spacetime as a gravitational condensate: USMEG-EFT emergence in comparison to Verlinde's entropic gravity
- Improving Signals, Declining Encoder: A Known-Answer Test of Label-Free Monitoring in Self-Supervised Learning
- Lyman Break Galaxy selection and redshift measurement with supervised contrastive learning
BibTeX
@misc{edwards2026widthmatched,
title = {{A Width-Matched Comparison of Hybrid Quantum-Classical Self-Supervised Learning for Fingerprint Recognition}},
author = {Maria S. Edwards and Kidwell Dlamini and Pin-An Lin and Wen-Hsien Hsu and Wen-Chieh Fang},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2609.39172},
eprint = {2609.39172},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.39172},
}
RIS
TY - UNPB
TI - A Width-Matched Comparison of Hybrid Quantum-Classical Self-Supervised Learning for Fingerprint Recognition
AU - Maria S. Edwards
AU - Kidwell Dlamini
AU - Pin-An Lin
AU - Wen-Hsien Hsu
AU - Wen-Chieh Fang
PY - 2026
DA - 2026/10/02
PB - arXiv
DO - 10.48550/arxiv.2609.39172
UR - https://arxiv.org/abs/2609.39172
AB - Fingerprint recognition is a widely deployed biometric, but supervised training requires large labeled enrollment sets. Self-supervised learning (SSL) removes this requirement, and hybrid quantum-classical models have been proposed to enrich the learned representations. Prior quantum SSL studies consider a single contrastive objective, so it is unclear whether reported benefits depend on the objective or can be attributed to the quantum circuit. We insert the QuFeX quantum feature-extraction module into three SSL frameworks, the contrastive SimCLR and MoCo v2 and the non-contrastive BYOL, and compare each hybrid with its classical counterpart at matched representation width (8 features, equal to 8 qubits) on the SOCOFing fingerprint dataset, with a CIFAR-10 control, using k-nearest-neighbor identification on encoder features. In single-run experiments the hybrid scores clearly higher for both contrastive objectives, whereas for BYOL a multi-seed analysis shows no reliable difference, suggesting that any benefit depends on the SSL objective. A hardware-efficient circuit (QNet) does not show the same gain. We examine whether the gains can be attributed to the quantum circuit, considering circuit architecture, trainable parameter count, nonlinearity, and the classical simulability of 8-qubit circuits.
ER -