Ten Years of Generative Adversarial Nets (GANs): A survey of the state-of-the-art

TL;DR

This survey analyzes a decade of Generative Adversarial Networks (GANs) research, focusing on architecture, training dynamics, evaluation, and applications. It presents the standard two-network framework with generator $G$ and discriminator $D$, whose minimax objective aims to approximate $p_{ ext{data}}$ using latent samples from $p_z$ via $ \\min_G \\max_D \\mathbb{E}_{x\sim p_{ ext{data}}}[\log D(x)] + \\mathbb{E}_{z\sim p_z}[\log(1 - D(G(z)))]$. The paper surveys a wide range of GAN variants (e.g., CGAN, CycleGAN, WGAN, StyleGAN) and their domains, theoretical insights linking adversarial training to divergences, and evaluation measures like Inception Score and Fréchet Inception Distance. It also discusses training challenges (mode collapse, vanishing gradients) and future directions that integrate GANs with Transformers, PINNs, LLMs, and Diffusion models, highlighting ethical and real-time considerations.

Abstract

Since their inception in 2014, Generative Adversarial Networks (GANs) have rapidly emerged as powerful tools for generating realistic and diverse data across various domains, including computer vision and other applied areas. Consisting of a discriminative network and a generative network engaged in a Minimax game, GANs have revolutionized the field of generative modeling. In February 2018, GAN secured the leading spot on the ``Top Ten Global Breakthrough Technologies List'' issued by the Massachusetts Science and Technology Review. Over the years, numerous advancements have been proposed, leading to a rich array of GAN variants, such as conditional GAN, Wasserstein GAN, CycleGAN, and StyleGAN, among many others. This survey aims to provide a general overview of GANs, summarizing the latent architecture, validation metrics, and application areas of the most widely recognized variants. We also delve into recent theoretical developments, exploring the profound connection between the adversarial principle underlying GAN and Jensen-Shannon divergence, while discussing the optimality characteristics of the GAN framework. The efficiency of GAN variants and their model architectures will be evaluated along with training obstacles as well as training solutions. In addition, a detailed discussion will be provided, examining the integration of GANs with newly developed deep learning frameworks such as Transformers, Physics-Informed Neural Networks, Large Language models, and Diffusion models. Finally, we reveal several issues as well as future research outlines in this field.

Figures (3)

• Figure 1: Architecture of GANs and its primary functions. In this example, different analytical tasks of GANs are categorized into synthetic data generation, style transfer, data augmentation, and anomaly detection.
• Figure 2: Timeline of the application-based GAN architectures reviewed in this study
• Figure 3: Diverse Applications of Generative Adversarial Networks (GANs) in various applied domains.