Self-Diagnosing GAN: Diagnosing Underrepresented Samples in Generative Adversarial Networks
arXiv:2102.12033
Abstract
Despite remarkable performance in producing realistic samples, Generative Adversarial Networks (GANs) often produce low-quality samples near low-density regions of the data manifold, e.g., samples of minor groups. Many techniques have been developed to improve the quality of generated samples, either by post-processing generated samples or by pre-processing the empirical data distribution, but at the cost of reduced diversity. To promote diversity in sample generation without degrading the overall quality, we propose a simple yet effective method to diagnose and emphasize underrepresented samples during training of a GAN. The main idea is to use the statistics of the discrepancy between the data distribution and the model distribution at each data instance. Based on the observation that the underrepresented samples have a high average discrepancy or high variability in discrepancy, we propose a method to emphasize those samples during training of a GAN. Our experimental results demonstrate that the proposed method improves GAN performance on various datasets, and it is especially effective in improving the quality and diversity of sample generation for minor groups.
arXiv commentAccepted to NeurIPS 2021
References in corpus (20)
- Conditional Generative Adversarial Nets
- GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
- Large Scale GAN Training for High Fidelity Natural Image Synthesis
- Spectral Normalization for Generative Adversarial Networks
- Which Training Methods for GANs do actually Converge?
- BEGAN: Boundary Equilibrium Generative Adversarial Networks
- VEEGAN: Reducing Mode Collapse in GANs using Implicit Variational Learning
- Unrolled Generative Adversarial Networks
- Making Convolutional Networks Shift-Invariant Again
- Deep multi-scale video prediction beyond mean square error
- Variational Approaches for Auto-Encoding Generative Adversarial Networks
- Banach Wasserstein GAN
- Bias Correction of Learned Generative Models using Likelihood-Free Importance Weighting
- LOGAN: Latent Optimisation for Generative Adversarial Networks
- Self-supervised GAN: Analysis and Improvement with Multi-class Minimax Game
- Discriminator Rejection Sampling
- Subsampling Generative Adversarial Networks: Density Ratio Estimation in Feature Space with Softplus Loss
- Mimicry: Towards the Reproducibility of GAN Research
- When Do Curricula Work?
- Mining GOLD Samples for Conditional GANs