生成对抗网络GAN综述

Generative Adversarial Networks GAN Overview

生成对抗网络(GAN)作为一种新的无监督学习算法框架得到越来越多研究者的青睐,已然成为当下的一个研究热点。GAN受启发于博弈论中的二人零和博弈理论,其独特的对抗训练思想能生成高质量的样本,具有比传统机器学习算法更加强大的特征学习和特征表达能力。目前GAN在计算机视觉领域尤其是在样本生成领域取得显著成功,每年有大量GAN相关研究的论文产出。针对GAN这一热点模型,首先介绍了GAN的研究现状;接着介绍了GAN的理论、框架,详细分析了GAN在训练过程中存在梯度消失和模式崩溃的原因;然后讨论了一些典型的GAN的改进模型,总结了它们理论的改进之处、优点、局限性、应用场景以及实现成本,同时还将GAN与VAE、RBM模型进行比较,总结出GAN的优势和劣势;最后展示了GAN在数据生成、图像超分辨率、图像风格转换等方面的应用成果,并探讨了GAN目前面临的挑战以及未来的研究方向。

As a new unsupervised learning algorithm framework, generative adversarial networks(GAN) has been favored by more and more researchers, and it has become a research hotspot. GAN is inspired by the two-person zero-sum game theory in game theory. Its unique confrontation training idea can generate high-quality samples and has more powerful feature learning and feature expression ability than traditional machine learning algorithms. At present, GAN has achieved remarkable success in the field of computer vision, especially in the field of sample generation. Each year, there are a large number of GAN-related research papers. For the hotspot model of GAN,firstly this paper introduces the research status of GAN;then introduces the theory and framework of GAN, and analyzes the reasons why GAN has gradient disappearance and mode collapse during training;then discusses some typical models of GAN. This paper summarizes the improvement, advantages, limitations, application scenarios and implementation costs of the theory. At the same time, this paper compares GAN with VAE(variational autoencoder)and RBM(restricted Boltzmann machine) models, and summarizes the advantages and disadvantages of GAN.Finally, the application results of GAN in data generation, image super-resolution, image style conversion, etc. are presented, and the challenges and future research directions of GAN are discussed.