Data-Efficient Instance Generation from Instance Discrimination

06/08/2021
by   Ceyuan Yang, et al.
2

Generative Adversarial Networks (GANs) have significantly advanced image synthesis, however, the synthesis quality drops significantly given a limited amount of training data. To improve the data efficiency of GAN training, prior work typically employs data augmentation to mitigate the overfitting of the discriminator yet still learn the discriminator with a bi-classification (i.e., real vs. fake) task. In this work, we propose a data-efficient Instance Generation (InsGen) method based on instance discrimination. Concretely, besides differentiating the real domain from the fake domain, the discriminator is required to distinguish every individual image, no matter it comes from the training set or from the generator. In this way, the discriminator can benefit from the infinite synthesized samples for training, alleviating the overfitting problem caused by insufficient training data. A noise perturbation strategy is further introduced to improve its discriminative power. Meanwhile, the learned instance discrimination capability from the discriminator is in turn exploited to encourage the generator for diverse generation. Extensive experiments demonstrate the effectiveness of our method on a variety of datasets and training settings. Noticeably, on the setting of 2K training images from the FFHQ dataset, we outperform the state-of-the-art approach with 23.5 improvement.

READ FULL TEXT

page 7

page 10

research
11/12/2021

Deceive D: Adaptive Pseudo Augmentation for GAN Training with Limited Data

Generative adversarial networks (GANs) typically require ample data for ...
research
04/07/2021

OpenGAN: Open-Set Recognition via Open Data Generation

Real-world machine learning systems need to analyze novel testing data t...
research
05/31/2021

OCT-GAN: Neural ODE-based Conditional Tabular GANs

Synthesizing tabular data is attracting much attention these days for va...
research
01/22/2019

A New CGAN Technique for Constrained Topology Design Optimization

This paper presents a new conditional GAN (named convex relaxing CGAN or...
research
09/20/2022

Improving GANs with A Dynamic Discriminator

Discriminator plays a vital role in training generative adversarial netw...
research
01/10/2020

microbatchGAN: Stimulating Diversity with Multi-Adversarial Discrimination

We propose to tackle the mode collapse problem in generative adversarial...
research
10/04/2021

GenCo: Generative Co-training on Data-Limited Image Generation

Training effective Generative Adversarial Networks (GANs) requires large...

Please sign up or login with your details

Forgot password? Click here to reset