AI Insight
This study compares hybrid quantum-classical neural networks with classical networks for fingerprint recognition using self-supervised learning approaches. The researchers tested a quantum feature extraction module (QuFeX) with three different self-supervised learning frameworks on fingerprint data, using networks matched at 8 features/qubits. Results showed quantum models performed better with contrastive learning methods (SimCLR and MoCo v2) but not with non-contrastive learning (BYOL), suggesting the benefit depends on the specific learning objective rather than quantum computation alone.
Why it matters
This research addresses whether quantum computing actually provides advantages for biometric recognition systems or if reported benefits are artifacts of experimental design. The findings suggest quantum approaches may only help with certain training methods, which is important for determining where to invest resources in quantum machine learning development for practical security applications.
Understand the Science
⚠️ Preprint – Noch nicht peer-reviewed
Dieser Artikel wurde noch nicht von unabhängigen Experten begutachtet. Die Ergebnisse sind vorläufig und sollten mit Vorsicht interpretiert werden.
-cross
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.