A Systematic Survey of Regularization and Normalization in GANs
arXiv:2008.08930 · doi:10.1145/3569928
Abstract
Generative Adversarial Networks (GANs) have been widely applied in different scenarios thanks to the development of deep neural networks. The original GAN was proposed based on the non-parametric assumption of the infinite capacity of networks. However, it is still unknown whether GANs can fit the target distribution without any prior information. Due to the overconfident assumption, many issues remain unaddressed in GANs' training, such as non-convergence, mode collapses, gradient vanishing. Regularization and normalization are common methods of introducing prior information to stabilize training and improve discrimination. Although a handful number of regularization and normalization methods have been proposed for GANs, to the best of our knowledge, there exists no comprehensive survey that primarily focuses on objectives and development of these methods, apart from some in-comprehensive and limited scope studies. In this work, we conduct a comprehensive survey on the regularization and normalization techniques from different perspectives of GANs training. First, we systematically describe different perspectives of GANs training and thus obtain the different objectives of regularization and normalization. Based on these objectives, we propose a new taxonomy. Furthermore, we compare the performance of the mainstream methods on different datasets and investigate the applications of regularization and normalization techniques that have been frequently employed in state-of-the-art GANs. Finally, we highlight potential future directions of research in this domain. Code and studies related to the regularization and normalization of GANs in this work is summarized on https://github.com/iceli1007/GANs-Regularization-Review.
Accepted to Acm Computing Survey
References in corpus (33)
- Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
- GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
- Improved Regularization of Convolutional Neural Networks with Cutout
- Diffusion Models Beat GANs on Image Synthesis
- Towards Principled Methods for Training Generative Adversarial Networks
- Neural Photo Editing with Introspective Adversarial Networks
- Regularization for Deep Learning: A Taxonomy
- Large Scale Image Completion via Co-Modulated Generative Adversarial Networks
- A survey on text generation using generative adversarial networks
- Generative Adversarial Networks for Spatio-temporal Data: A Survey
- Diversity-Sensitive Conditional Generative Adversarial Networks
- Image Augmentations for GAN Training
- Improving Generalization and Stability of Generative Adversarial Networks
- Diffusion-GAN: Training GANs with Diffusion
- CSCNet: Contextual Semantic Consistency Network for Trajectory Prediction in Crowded Spaces
- Consistency Regularization for Generative Adversarial Networks
- Lipschitz Generative Adversarial Nets
- Generative Adversarial Networks for Image Super-Resolution: A Survey
- Data-Efficient Instance Generation from Instance Discrimination
- Deceive D: Adaptive Pseudo Augmentation for GAN Training with Limited Data
- GAN-QP: A Novel GAN Framework without Gradient Vanishing and Lipschitz Constraint
- Regularization Methods for Generative Adversarial Networks: An Overview of Recent Studies
- A Comprehensive Survey on Data-Efficient GANs in Image Generation
- On Solving Minimax Optimization Locally: A Follow-the-Ridge Approach
- How Well Do WGANs Estimate the Wasserstein Metric?
- Real or Not Real, that is the Question
- Wasserstein GANs Work Because They Fail (to Approximate the Wasserstein Distance)
- SSD-GAN: Measuring the Realness in the Spatial and Spectral Domains
- Training Generative Adversarial Networks by Solving Ordinary Differential Equations
- Regularizing Generative Adversarial Networks under Limited Data
- Towards Efficient and Unbiased Implementation of Lipschitz Continuity in GANs
- View Vertically: A Hierarchical Network for Trajectory Prediction via Fourier Spectrums
- FakeCLR: Exploring Contrastive Learning for Solving Latent Discontinuity in Data-Efficient GANs