Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

A Lightweight Hybrid Quantum–Classical Neural Architecturewith Structured Representation Adaptation

Wan Nursyafeeza Binti Wan Mohd Nasir, Sheikh Sayed Bin Rahman, Chimeremma Sandra Amadi, Taesoo Jun

Published Jan 1, 2026• 0 citations
Publisher
DOI
10.2139/ssrn.6946184

Related papers

Domain adaptation based on hybrid classical-quantum neural network

Yihua Wu, Chunhui Wu, Anqi Zhang +1 more

Quantum Information Processing•Jun 26, 2023•1 citation

Hybrid Quantum–Classical Convolutional Neural Network for Medical Image Classification

Geethanjali, Jithinkrishnan P G, Jaikrishnan M O +3 more

Significant advances have been made in quantum computing, enabling researchers to create hybrid quantum- classical learning models for various tasks, including medical image classification. The classical approach to deep learning often faces challenges with limited data availability, high computational costs, and poor generalization. The hybrid Quantum- Classical Convolutional Neural Networks (HQCNNs) provide a new direction for overcoming the challenges faced by classical approaches. The hybrid approach enables the use of quantum circuits for feature encoding in higher-dimensional space, improving discrimination between patterns and reducing overfitting. In this paper, a hybrid quantum-classical convolutional neural network is developed for medical image classification. The hybrid approach is expected to capture non-linear relationships within medical image data while maintaining a lightweight network structure. The hybrid approach is found to have significant potential for efficient operation with limited data availability and noisy data. The classical approach is often outperformed by the hybrid approach. The focus of this paper is on architectural design and analysis of the developed hybrid quantum-classical convolutional neural network. The advantages of the developed hybrid quantum-classical convolutional neural network are de- lineated, including current challenges with quantum hardware.

International Research Journal on Advanced Engineering and Management (IRJAEM)•May 5, 2026•0 citations

A hybrid quantum-inspired neural networks with sequence inputs

Panchi Li, Hong Yu Xiao, Fuhua Shang +3 more

Neurocomputing•Mar 4, 2013•53 citations

A Novel Lightweight Hybrid MLP and Quantum Neural Network Model for Bit Error Rate Prediction in NOMA-Enabled Visible Light Communication Systems

Vipul Dixit, Bhuvan Unhelkar, Abhay Bansal +1 more

In this letter, a lightweight hybrid Multi-Layer Perceptron-Quantum Neural Network (MLP-QNN) architecture for Bit Error Rate (BER) prediction in Non-Orthogonal Multiple Access-Visible Light Communication (NOMA-VLC) systems is proposed. The proposed approach integrates classical preprocessing with variational quantum circuits (VQC) using ZFeatureMap encoding and TwoLocal ansatz in a compact 170-parameter structure implemented through Qiskit-PyTorch framework. The quantum circuits serve as enhanced feature extractors to capture complex NOMA-VLC dynamics. Simulation results demonstrate the superiority of the hybrid quantum-classical approach, achieving R2scores of 0.9989 and 0.9985 for far and near users, with 1 – 1.5dBSNR gains over classical MLP, advanced classical and analytical methods at reduced computational requirements. The model achieves the high accuracy and ≈0.8dBSNR gains under noisy simulation, confirming quantum noise robustness.

IEEE Wireless Communications Letters•Jan 1, 2026•1 citation

Architecture of Quantum Neural Networks: Design and Implementation

Fathima Nihla Latheef, Rubell Marion Lincy G

Springer eBooks•Jan 1, 2025•0 citations

Quantum-Inspired Fine-Tuning for Few-Shot AIGC Detection via Phase-Structured Reparameterization

Kaiyang Xing, Han Fang, Zhaoyun Chen +4 more

Recent studies show that quantum neural networks (QNNs) generalize well in few-shot regimes. To extend this advantage to large-scale tasks, we propose Q-LoRA, a quantum-enhanced fine-tuning scheme that integrates lightweight QNNs into the low-rank adaptation (LoRA) adapter. Applied to AI-generated content (AIGC) detection, Q-LoRA consistently outperforms standard LoRA under few-shot settings. We analyze the source of this improvement and identify two possible structural inductive biases from QNNs: (i) phase-aware representations, which encode richer information across orthogonal amplitude-phase components, and (ii) norm-constrained transformations, which stabilize optimization via inherent orthogonality. However, Q-LoRA incurs non-trivial overhead due to quantum simulation. Motivated by our analysis, we further introduce H-LoRA, a fully classical variant that applies the Hilbert transform within the LoRA adapter to retain similar phase structure and constraints. Experiments on few-shot AIGC detection show that both Q-LoRA and H-LoRA outperform standard LoRA by over 5% accuracy, with H-LoRA achieving comparable accuracy at significantly lower cost in this task.

cs.LGCross-listedarXiv:2603.02281•Mar 3, 2026•1 citation
Significant advances have been made in quantum computing, enabling researchers to create hybrid quantum- classical learning models for various tasks, including medical image classification. The classical approach to deep learning often faces challenges with limited data availability, high computational costs, and poor generalization. The hybrid Quantum- Classical Convolutional Neural Networks (HQCNNs) provide a new direction for overcoming the challenges faced by classical approaches. The hybrid approach enables the use of quantum circuits for feature encoding in higher-dimensional space, improving discrimination between patterns and reducing overfitting. In this paper, a hybrid quantum-classical convolutional neural network is developed for medical image classification. The hybrid approach is expected to capture non-linear relationships within medical image data while maintaining a lightweight network structure. The hybrid approach is found to have significant potential for efficient operation with limited data availability and noisy data. The classical approach is often outperformed by the hybrid approach. The focus of this paper is on architectural design and analysis of the developed hybrid quantum-classical convolutional neural network. The advantages of the developed hybrid quantum-classical convolutional neural network are de- lineated, including current challenges with quantum hardware.
In this letter, a lightweight hybrid Multi-Layer Perceptron-Quantum Neural Network (MLP-QNN) architecture for Bit Error Rate (BER) prediction in Non-Orthogonal Multiple Access-Visible Light Communication (NOMA-VLC) systems is proposed. The proposed approach integrates classical preprocessing with variational quantum circuits (VQC) using ZFeatureMap encoding and TwoLocal ansatz in a compact 170-parameter structure implemented through Qiskit-PyTorch framework. The quantum circuits serve as enhanced feature extractors to capture complex NOMA-VLC dynamics. Simulation results demonstrate the superiority of the hybrid quantum-classical approach, achieving R2scores of 0.9989 and 0.9985 for far and near users, with 1 – 1.5dBSNR gains over classical MLP, advanced classical and analytical methods at reduced computational requirements. The model achieves the high accuracy and ≈0.8dBSNR gains under noisy simulation, confirming quantum noise robustness.
Recent studies show that quantum neural networks (QNNs) generalize well in few-shot regimes. To extend this advantage to large-scale tasks, we propose Q-LoRA, a quantum-enhanced fine-tuning scheme that integrates lightweight QNNs into the low-rank adaptation (LoRA) adapter. Applied to AI-generated content (AIGC) detection, Q-LoRA consistently outperforms standard LoRA under few-shot settings. We analyze the source of this improvement and identify two possible structural inductive biases from QNNs: (i) phase-aware representations, which encode richer information across orthogonal amplitude-phase components, and (ii) norm-constrained transformations, which stabilize optimization via inherent orthogonality. However, Q-LoRA incurs non-trivial overhead due to quantum simulation. Motivated by our analysis, we further introduce H-LoRA, a fully classical variant that applies the Hilbert transform within the LoRA adapter to retain similar phase structure and constraints. Experiments on few-shot AIGC detection show that both Q-LoRA and H-LoRA outperform standard LoRA by over 5% accuracy, with H-LoRA achieving comparable accuracy at significantly lower cost in this task.