Self-Diagnosing GAN: Diagnosing Underrepresented Samples in Generative Adversarial Networks

02/24/2021
by   Jinhee Lee, et al.
17

Despite remarkable performance in producing realistic samples, Generative Adversarial Networks (GANs) often produce low-quality samples near low-density regions of the data manifold. Recently, many techniques have been developed to improve the quality of generated samples, either by rejecting low-quality samples after training or by pre-processing the empirical data distribution before training, but at the cost of reduced diversity. To guarantee both the quality and the diversity, 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 of generated samples with minor features.

READ FULL TEXT

page 17

page 18

page 19

page 20

page 21

page 23

page 24

page 25

research
10/21/2019

Mining GOLD Samples for Conditional GANs

Conditional generative adversarial networks (cGANs) have gained a consid...
research
07/30/2020

Instance Selection for GANs

Recent advances in Generative Adversarial Networks (GANs) have led to th...
research
11/14/2016

Generative Models and Model Criticism via Optimized Maximum Mean Discrepancy

We propose a method to optimize the representation and distinguishabilit...
research
10/15/2021

MaGNET: Uniform Sampling from Deep Generative Network Manifolds Without Retraining

Deep Generative Networks (DGNs) are extensively employed in Generative A...
research
06/18/2020

MMCGAN: Generative Adversarial Network with Explicit Manifold Prior

Generative Adversarial Network(GAN) provides a good generative framework...
research
11/04/2019

Improved BiGAN training with marginal likelihood equalization

We propose a novel training procedure for improving the performance of g...
research
10/12/2022

What can we learn about a generated image corrupting its latent representation?

Generative adversarial networks (GANs) offer an effective solution to th...

Please sign up or login with your details

Forgot password? Click here to reset