Instance Selection for GANs

by   Terrance DeVries, et al.

Recent advances in Generative Adversarial Networks (GANs) have led to their widespread adoption for the purposes of generating high quality synthetic imagery. While capable of generating photo-realistic images, these models often produce unrealistic samples which fall outside of the data manifold. Several recently proposed techniques attempt to avoid spurious samples, either by rejecting them after generation, or by truncating the model's latent space. While effective, these methods are inefficient, as large portions of model capacity are dedicated towards representing samples that will ultimately go unused. In this work we propose a novel approach to improve sample quality: altering the training dataset via instance selection before model training has taken place. To this end, we embed data points into a perceptual feature space and use a simple density model to remove low density regions from the data manifold. By refining the empirical data distribution before training we redirect model capacity towards high-density regions, which ultimately improves sample fidelity. We evaluate our method by training a Self-Attention GAN on ImageNet at 64x64 resolution, where we outperform the current state-of-the-art models on this task while using 1/2 of the parameters. We also highlight training time savings by training a BigGAN on ImageNet at 128x128 resolution, achieving a 66 the baseline model with less than 1/4 the training time.


page 4

page 7

page 8

page 13

page 14

page 15

page 16


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

Despite remarkable performance in producing realistic samples, Generativ...

AE-OT-GAN: Training GANs from data specific latent distribution

Though generative adversarial networks (GANs) areprominent models to gen...

Generating High Fidelity Data from Low-density Regions using Diffusion Models

Our work focuses on addressing sample deficiency from low-density region...

LatentAugment: Data Augmentation via Guided Manipulation of GAN's Latent Space

Data Augmentation (DA) is a technique to increase the quantity and diver...

On the "steerability" of generative adversarial networks

An open secret in contemporary machine learning is that many models work...

Instance-Conditioned GAN

Generative Adversarial Networks (GANs) can generate near photo realistic...

Please sign up or login with your details

Forgot password? Click here to reset