Improving Generalization and Stability of Generative Adversarial Networks

2019

Citations in the corpus

• Catastrophic forgetting and mode collapse in GANs (2020, neutral): “3).5 GAN-0GP, GAN-R1, and WGAN-GP exhibit better convergence behaviors: for many images, the digits stay the same during training.”
• On catastrophic forgetting and mode collapse in Generative Adversarial Networks (2018, unknown): “Section V explains how state-of-the-art methods for stabilizing GANs such as Wasserstein GAN [7, 8], zero-centered gradient penalty on training examples (GAN-R1) [9], zerocentered gradient penalty on interpolated samples (GAN-0GP) [10], and optimizers with momentum, can prevent CF and mode collapse.”
• Unsupervised anomaly localization using VAE and beta-VAE (2020, unknown): “Recent works such as [20] propose substantial upgrades, however other works such as [17] still supports that GANs have more trouble than other generative models to cover the whole distribution support.”
• Quality Aware Generative Adversarial Networks (2019, unknown): “Thanh-Tung et al. [11] propose another variant of the zero-centered gradient penalty for improving the convergence and generalizing capability of GANs.”
• DF-GAN: Deep Fusion Generative Adversarial Networks for Text-to-Image Synthesis (2020, unknown): “For solving this problem, some researches propose the Gradient Penalty [44], [45], [46], [35] and Spectral Normalization [40] which frees WGANs from this challenging constraint in unconditional generation tasks.”