Abstract Quantum generative adversarial networks (QGANs) can effectively enhance the performance and efficiency of classical GANs by utilizing the parallelism of quantum computation and quantum superposition. However, QGANs typically suffer from mode collapse during the training process of generative tasks. It would make the generator only be able to generate partially correct data approximately. To solve this problem, an unrolled QGAN model based on a hybrid quantum-classical framework is constructed. The unrolled QGAN can match the generator with a better discriminator by separately training the discriminator prior to the training on the generator. The model is applied to generate quantum and Gaussian distributions, and comparative experiments are performed between the QGAN and the proposed unrolled one. Mean value, KL divergence, and standard deviation are calculated and compared to evaluate the generative performance of the model. Numerical and experimental results show that the proposed unrolled QGAN can increase the diversity and coverage of generated data distributions, significantly enhancing the generative effect.
Abstract Quantum generative adversarial networks (QGANs) can effectively enhance the performance and efficiency of classical GANs by utilizing the parallelism of quantum computation and quantum superposition. However, QGANs typically suffer from mode collapse during the training process of generative tasks. It would make the generator only be able to generate partially correct data approximately. To solve this problem, an unrolled QGAN model based on a hybrid quantum-classical framework is constructed. The unrolled QGAN can match the generator with a better discriminator by separately training the discriminator prior to the training on the generator. The model is applied to generate quantum and Gaussian distributions, and comparative experiments are performed between the QGAN and the proposed unrolled one. Mean value, KL divergence, and standard deviation are calculated and compared to evaluate the generative performance of the model. Numerical and experimental results show that the proposed unrolled QGAN can increase the diversity and coverage of generated data distributions, significantly enhancing the generative effect.
Synthetic biological data play a crucial role in bioinformatics research, particularly in scenarios involving limited data availability and strict privacy constraints. However, conventional generative adversarial networks often face challenges in simultaneously preserving local statistical characteristics and global data distributions, motivating the exploration of quantum-enhanced generative models. This paper proposes and evaluates a hybrid quantum–classical generative adversarial network (QGAN) for efficient and high-quality biological data generation. The generator is implemented using a variational quantum circuit that encodes classical biological features into parameterized quantum states, while the discriminator is realized through a classical neural network architecture. The proposed framework is evaluated against classical GAN models using quantitative metrics, including the Classifier Two-Sample Test (C2ST), Maximum Mean Discrepancy (MMD) and Wasserstein Distance (WD). Experimental results show that the QGAN achieves stronger local statistical distribution matching, reflected by lower MMD values, whereas the Classical GAN demonstrates more stable global distribution alignment through lower WD. Furthermore, the analysis reveals that the stochastic sampling behavior of parameterized quantum circuits contributes to higher output variance during training, influencing large-scale distribution consistency. These findings indicate that hybrid quantum–classical generative models can effectively preserve localized statistical structures while highlighting the trade-off between local fidelity and global distribution stability, supporting their potential for future applications in genomics, drug discovery, and privacy-preserving biomedical data synthesis.
This paper presents a novel framework for Quantum Generative Adversarial Networks (QGANs) that integrates Quantum Kernel Learning into the discriminator architecture. By replacing the classical discriminator with a Quantum Kernel Discriminator (QKD), the proposed model leverages high-dimensional Hilbert space mappings to enhance separability between real and generated data. The quantum discriminator employs Pauli Feature Maps and Fidelity Quantum Kernels to evaluate sample similarity in the quantum space, leading to improved training stability and reduced mode collapse. Empirical results on the MNIST dataset demonstrate faster convergence, higher fooling scores, and improved sample diversity compared to classical GANs. Additionally, the application of Quantum Kernel Methods to the Iris dataset highlights the generalization capabilities of the QKD approach. These findings establish the potential of hybrid quantum-classical architectures in advancing generative modeling and open new avenues for applying quantum machine learning in low-data and high-complexity domains.
This paper investigates the application of Quantum Generative Adversarial Networks (QGANs) for the generation of synthetic data. Traditional Generative Adversarial Networks (GANs) often face challenges in producing high-quality synthetic data, especially when dealing with complex data distributions. We propose a novel framework utilizing quantum circuits to construct both the generator and discriminator within a QGAN architecture. The core innovation lies in harnessing quantum superposition and interference during the adversarial training process to potentially improve the realism and diversity of the generated synthetic data. Our approach introduces a theoretical foundation and a conceptual design for QGANs, paving the way for future research and development in this emerging field. The primary goal is to demonstrate the potential benefits of incorporating quantum mechanical principles into the GAN framework, leading to enhanced synthetic data generation capabilities. The performance of a QGAN is expected to be superior to a classical GAN due to the unique properties of quantum systems.
One of the most promising applications in the era of Noisy Intermediate-Scale Quantum (NISQ) computing is quantum generative adversarial networks (QGANs), which offer significant quantum advantages over classical machine learning in various domains. However, QGANs suffer from mode collapse and lack explicit control over the features of generated outputs. To overcome these limitations, we propose InfoQGAN, a novel quantum-classical hybrid generative adversarial network that integrates the principles of InfoGAN with a QGAN architecture. Our approach employs a variational quantum circuit for data generation, a classical discriminator, and a Mutual Information Neural Estimator (MINE) to explicitly optimize the mutual information between latent codes and generated samples. Numerical simulations on synthetic 2D distributions and Iris dataset augmentation demonstrate that InfoQGAN effectively mitigates mode collapse while achieving robust feature disentanglement in the quantum generator. By leveraging these advantages, InfoQGAN not only enhances training stability but also improves data augmentation performance through controlled feature generation. These results highlight the potential of InfoQGAN as a foundational approach for advancing quantum generative modeling in the NISQ era.
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.
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.