Review of Quantum Generative Adversarial Networks

Quantum Generative Adversarial Networks (QGANs) combine the generative capabilities of classical GANs with the high-dimensional representation power of quantum computing, aiming to improve generation quality and computational efficiency. Since their introduction in 2018, QGANs have evolved into various architectures, depending on the quantum or classical nature of the generator and discriminator. Fully quantum models and hybrid architectures with quantum generators have become key research directions, with notable progress in model design and training stability. This paper provides a comprehensive review of QGANs, covering their fundamental principles, major model types, representative advancements, and practical applications. Additionally, it discusses current challenges and explores future directions in the pursuit of quantum advantage and real-world deployment.