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

Radial Basis Function-based Quantum Hybrid Classical Generative Adversarial Networks for Enhanced Image Quality and Training Stability

Zuyu Xu, Tao Yang, Pengnian Cai, Kang Shen, Yuanming Hu, Bin Lv, Shixian Chen, Yunlai Zhu, Zuheng Wu, Jun Wang, Yuehua Dai

Published Apr 4, 2024• 2 citations
Publisher

Abstract

Abstract Quantum Generative Adversarial Networks (QGANs), as the quantum version to classical Generative Adversarial Networks, exhibit exponential advantages in certain aspects, garnering considerable attention. However, within this nascent field, challenges persist in the synthesis of image quality and the stability of training in QGANs. In this work, we introduce a Hybrid Quantum Classical Generative Adversarial Network (HQCGAN), incorporating a classical discriminator constructed using Radial Basis Function Neural Networks (RBFNN). Harnessing the superior non-linear data processing capabilities and inherent resilience to image noise of RBFNNs, our HQCGAN significantly enhances its proficiency in generating high-fidelity grayscale images characterized by discrete value distributions. Through a series of meticulous experiments that evaluated the training cross-validation scores and the robustness of the loss functions, we have demonstrated the exceptional performance of our HQCGAN model, especially in the presence of noisy input data. These findings contribute meaningfully to the burgeoning field of quantum generative models, underscoring the vital role played by classical machine learning components in augmenting the overall efficacy of quantum approaches. The incorporation of RBFNNs within a quantum framework in our study offers novel perspectives to address prevailing challenges related to image quality and training stability, marking a substantial progression in the evolution of quantum generative adversarial networks.

DOI
10.21203/rs.3.rs-4195599/v1

Related papers

HQSP-GAN: A hybrid quantum–classical architecture for high-quality image generation

Minghua Pan, Bing Hua Huang, Shaohua Sun +3 more

Abstract Quantum machine learning has attracted considerable attention due to its potential in classification and recognition tasks. As a significant branch of this field, quantum generative adversarial networks (QGANs) have achieved notable progress in image generation. Although existing models have improved in terms of generation quality and efficiency, their performance remains suboptimal. To overcome these limitations, we propose HQSP-GAN, a hybrid quantum–classical generative adversarial model integrating a shared parameter mechanism. This model innovatively designs a quantum parameter sharing scheme to reduce the number of trainable parameters in the quantum module. Additionally, by incorporating attention mechanisms, the Wasserstein distance, and gradient penalty techniques, HQSP-GAN enhances training stability and generation performance. Extensive experiments on standard datasets such as MNIST and Fashion MNIST demonstrate that HQSP-GAN can generate high-quality grayscale images without additional dimensionality reduction or preprocessing. Moreover, it outperforms SOTA models including QINR-GAN and PQWGAN in metrics such as SSIM and PSNR, highlighting the efficiency and practicality of its architectural design.

Chinese Physics B•Nov 17, 2025•0 citations

QGAN Driven Hybrid Quantum—Classical Architectures for Efficient and Noise Resilient Machine Learning

B Gouthami, C. Janani, R.Banu priya +3 more

This paper introduces a novel hybrid quantumclassical model using QGAN (quantum generative adversarial network) to achieve more efficient, convergent and noiserobust machine learning for nearterm quantum (NISQ) applications. Unlike the existing GAN type models that were purely classical or purely quantum, to propose a quantum generator consisting of shallow variation circuits (48 qubits) together with a classical discriminator that is crucial for the stability of adversarial training. The empirical results on the MNIST dataset showed that were able to perform classification with an accuracy of 94.2%, an improvement of over 3.2% from a straightforward classical only approach, while NVGAN also improved its generative performance to a Fréchet Inception Distance (FID) score of 38.7 or 26% in quality enhancement from synthesis perspective. Both trainings converged 50 epochs earlier than each training when used the hybrid model (30% improvement in convergence). The key innovation of this work is bringing (1) to bear both the utility of the quantumenhanced feature representation to have at disposal, and (2) classical stability measures and erroraware optimizations (such as parameter clipping, early stopping), such that its hybridization enabled it to continue operating with$\mathbf{> 9 0 \%}$accuracy under cases of reasonable depolarizing/amplitude damping noise on an MLP where purely quantum formulations decayed to (approx 70%). Altogether, the paper validates the presented hybrid QGAN approach as a statistically substantiated noise robust and scalable representational tool for real world applications of quantum enhanced machine learning tasks on NISQ platforms.

International Conference on Communication, Computing and Electronics Systems (ICCCES)•Jan 21, 2026•0 citations

Quantum generative adversarial networks with dual generators

Quangong Ma, Chaolong Hao, Nianwen Si +1 more

Abstract Quantum generative adversarial networks (QGANs) have demonstrated strong capabilities in tasks like synthetic data generation and detecting anomalies. Recent developments have increasingly integrated traditional machine learning techniques to boost the performance of QGANs. Motivated by this progress, we propose an innovative QGAN architecture that incorporates a classical learning component and employs a dual-generator design. Our approach improves upon the traditional hybrid quantum–classical GAN structure and introduces a redesigned loss function tailored for the new model. Experiments on multiple datasets indicate that our method surpasses previous techniques in image generation quality, achieving a 1.38% average reduction in Fréchet inception distance scores compared to the current state-of-the-art, and improvements of 6.52%, 0.36%, and 0.38% in structural similarity index, cosine similarity, and peak signal-to-noise ratio metrics, respectively. Additionally, our architecture supports the generation of larger images (up to 78 × 78 ), as verified on the CelebA dataset. Simulations conducted in noisy conditions further confirm the robustness and effectiveness of both the proposed architecture and loss function.

Machine Learning Science and Technology•Sep 25, 2025•1 citation

BAE-QCGAN: a hybrid quantum-classical generative adversarial network

Han Qi, Zhichao Liang, Abdullah Gani +2 more

With the speedy advancement of hardware, quantum machine learning has garnered widespread attention. Generative Adversarial Networks (GANs) are widely adopted quantum machine learning frameworks used for providing highstandard audio content, images, and videos. Utilizing quantum computing for image generation has been found to offer several potential advantages over non-quantum technologies. Current quantum Generative Adversarial Networks (QGANs) still face challenges, such as the inability to generate high-quality images. In the current investigation, we propose a QGAN model based on Autoencoders (BAE-QCGAN) and evaluate its performance. By drawing upon the entanglement and expression capabilities of quantum circuits, our composite system outperforms classical GANs in terms of Frechet Inception Distance, requiring fewer training parameters and fewer iterations to converge. We also demonstrate that BAEQCGAN outperforms another quantum GAN, the patch-GAN, and is adapted for existing Noisy Intermediate Scale Quantum (NISQ) computers.

International Conference on Physics and Quantum Computing (ICPQC)•Aug 5, 2025•0 citations

Quantum Generative Neural Networks for Imaging Applications

Sunil Vittal, David Ramirez, Glen S. Uehara +2 more

Quantum Machine Learning has received a lot of attention recently in the research community due to its potential to provide computational advantages in areas including training neural networks and big data analytics. In this paper, we focus on generative approaches using quantum machine learning. Specifically, we train QuantumClassical Hybrid Generative Adversarial Networks (qGANs) for imaging applications. We use the MNIST and Fashion-MNIST datasets in our qGAN simulations, and we present examples that generate small-scale grayscale images of the zero digit and shirts classes. In our research, we used Xanadu's PennyLane quantum programming software to implement and evaluate our algorithms. Challenges in our study included long computational times and large memory requirements for training the qGAN. However, our results revealed that our proposed hybrid model achieves a higher average PSNR, a lower FID score, but a lower Inception score than a classical GAN.

International Conference on Information, Intelligence, Systems & Applications (IISA)•Jul 10, 2025•2 citations

Quantum Generative Adversarial Networks (QGANs) for Data Augmentation

Jincheng Zhang

This paper investigates the application of Quantum Generative Adversarial Networks (QGANs) to data augmentation for machine learning. Traditional Generative Adversarial Networks (GANs) rely on classical neural networks, while this research explores leveraging quantum circuits as generators within a QGAN architecture. The core idea is to train a quantum generator to produce synthetic data samples, supplementing existing datasets with novel instances. The discriminator remains a classical neural network, tasked with distinguishing between real and generated data. We present a theoretical framework for this approach, outlining the mathematical formulation and key components. Specifically, we introduce the notation (x, s_1, s_2, ..., s_K) to represent data samples, generator states, and relevant parameters. The objective function for the QGAN is defined as minimizing the Jensen-Shannon divergence between the real and generated data distributions. The effectiveness of this approach is discussed, highlighting potential advantages and challenges. This work contributes to the emerging field of quantum machine learning and offers a novel strategy for improving the performance of machine learning models, particularly in scenarios with limited training data.

•Aug 31, 2026•0 citations
Abstract Quantum Generative Adversarial Networks (QGANs), as the quantum version to classical Generative Adversarial Networks, exhibit exponential advantages in certain aspects, garnering considerable attention. However, within this nascent field, challenges persist in the synthesis of image quality and the stability of training in QGANs. In this work, we introduce a Hybrid Quantum Classical Generative Adversarial Network (HQCGAN), incorporating a classical discriminator constructed using Radial Basis Function Neural Networks (RBFNN). Harnessing the superior non-linear data processing capabilities and inherent resilience to image noise of RBFNNs, our HQCGAN significantly enhances its proficiency in generating high-fidelity grayscale images characterized by discrete value distributions. Through a series of meticulous experiments that evaluated the training cross-validation scores and the robustness of the loss functions, we have demonstrated the exceptional performance of our HQCGAN model, especially in the presence of noisy input data. These findings contribute meaningfully to the burgeoning field of quantum generative models, underscoring the vital role played by classical machine learning components in augmenting the overall efficacy of quantum approaches. The incorporation of RBFNNs within a quantum framework in our study offers novel perspectives to address prevailing challenges related to image quality and training stability, marking a substantial progression in the evolution of quantum generative adversarial networks.
Abstract Quantum machine learning has attracted considerable attention due to its potential in classification and recognition tasks. As a significant branch of this field, quantum generative adversarial networks (QGANs) have achieved notable progress in image generation. Although existing models have improved in terms of generation quality and efficiency, their performance remains suboptimal. To overcome these limitations, we propose HQSP-GAN, a hybrid quantum–classical generative adversarial model integrating a shared parameter mechanism. This model innovatively designs a quantum parameter sharing scheme to reduce the number of trainable parameters in the quantum module. Additionally, by incorporating attention mechanisms, the Wasserstein distance, and gradient penalty techniques, HQSP-GAN enhances training stability and generation performance. Extensive experiments on standard datasets such as MNIST and Fashion MNIST demonstrate that HQSP-GAN can generate high-quality grayscale images without additional dimensionality reduction or preprocessing. Moreover, it outperforms SOTA models including QINR-GAN and PQWGAN in metrics such as SSIM and PSNR, highlighting the efficiency and practicality of its architectural design.
This paper introduces a novel hybrid quantumclassical model using QGAN (quantum generative adversarial network) to achieve more efficient, convergent and noiserobust machine learning for nearterm quantum (NISQ) applications. Unlike the existing GAN type models that were purely classical or purely quantum, to propose a quantum generator consisting of shallow variation circuits (48 qubits) together with a classical discriminator that is crucial for the stability of adversarial training. The empirical results on the MNIST dataset showed that were able to perform classification with an accuracy of 94.2%, an improvement of over 3.2% from a straightforward classical only approach, while NVGAN also improved its generative performance to a Fréchet Inception Distance (FID) score of 38.7 or 26% in quality enhancement from synthesis perspective. Both trainings converged 50 epochs earlier than each training when used the hybrid model (30% improvement in convergence). The key innovation of this work is bringing (1) to bear both the utility of the quantumenhanced feature representation to have at disposal, and (2) classical stability measures and erroraware optimizations (such as parameter clipping, early stopping), such that its hybridization enabled it to continue operating with>90%\mathbf{> 9 0 \%}>90%accuracy under cases of reasonable depolarizing/amplitude damping noise on an MLP where purely quantum formulations decayed to (approx 70%). Altogether, the paper validates the presented hybrid QGAN approach as a statistically substantiated noise robust and scalable representational tool for real world applications of quantum enhanced machine learning tasks on NISQ platforms.
Abstract Quantum generative adversarial networks (QGANs) have demonstrated strong capabilities in tasks like synthetic data generation and detecting anomalies. Recent developments have increasingly integrated traditional machine learning techniques to boost the performance of QGANs. Motivated by this progress, we propose an innovative QGAN architecture that incorporates a classical learning component and employs a dual-generator design. Our approach improves upon the traditional hybrid quantum–classical GAN structure and introduces a redesigned loss function tailored for the new model. Experiments on multiple datasets indicate that our method surpasses previous techniques in image generation quality, achieving a 1.38% average reduction in Fréchet inception distance scores compared to the current state-of-the-art, and improvements of 6.52%, 0.36%, and 0.38% in structural similarity index, cosine similarity, and peak signal-to-noise ratio metrics, respectively. Additionally, our architecture supports the generation of larger images (up to 78 × 78 ), as verified on the CelebA dataset. Simulations conducted in noisy conditions further confirm the robustness and effectiveness of both the proposed architecture and loss function.
With the speedy advancement of hardware, quantum machine learning has garnered widespread attention. Generative Adversarial Networks (GANs) are widely adopted quantum machine learning frameworks used for providing highstandard audio content, images, and videos. Utilizing quantum computing for image generation has been found to offer several potential advantages over non-quantum technologies. Current quantum Generative Adversarial Networks (QGANs) still face challenges, such as the inability to generate high-quality images. In the current investigation, we propose a QGAN model based on Autoencoders (BAE-QCGAN) and evaluate its performance. By drawing upon the entanglement and expression capabilities of quantum circuits, our composite system outperforms classical GANs in terms of Frechet Inception Distance, requiring fewer training parameters and fewer iterations to converge. We also demonstrate that BAEQCGAN outperforms another quantum GAN, the patch-GAN, and is adapted for existing Noisy Intermediate Scale Quantum (NISQ) computers.
Quantum Machine Learning has received a lot of attention recently in the research community due to its potential to provide computational advantages in areas including training neural networks and big data analytics. In this paper, we focus on generative approaches using quantum machine learning. Specifically, we train QuantumClassical Hybrid Generative Adversarial Networks (qGANs) for imaging applications. We use the MNIST and Fashion-MNIST datasets in our qGAN simulations, and we present examples that generate small-scale grayscale images of the zero digit and shirts classes. In our research, we used Xanadu's PennyLane quantum programming software to implement and evaluate our algorithms. Challenges in our study included long computational times and large memory requirements for training the qGAN. However, our results revealed that our proposed hybrid model achieves a higher average PSNR, a lower FID score, but a lower Inception score than a classical GAN.
This paper investigates the application of Quantum Generative Adversarial Networks (QGANs) to data augmentation for machine learning. Traditional Generative Adversarial Networks (GANs) rely on classical neural networks, while this research explores leveraging quantum circuits as generators within a QGAN architecture. The core idea is to train a quantum generator to produce synthetic data samples, supplementing existing datasets with novel instances. The discriminator remains a classical neural network, tasked with distinguishing between real and generated data. We present a theoretical framework for this approach, outlining the mathematical formulation and key components. Specifically, we introduce the notation (x, s_1, s_2, ..., s_K) to represent data samples, generator states, and relevant parameters. The objective function for the QGAN is defined as minimizing the Jensen-Shannon divergence between the real and generated data distributions. The effectiveness of this approach is discussed, highlighting potential advantages and challenges. This work contributes to the emerging field of quantum machine learning and offers a novel strategy for improving the performance of machine learning models, particularly in scenarios with limited training data.