MineGAN++: Mining Generative Models for Efficient Knowledge Transfer to Limited Data Domains

04/28/2021
by   Yaxing Wang, et al.
29

GANs largely increases the potential impact of generative models. Therefore, we propose a novel knowledge transfer method for generative models based on mining the knowledge that is most beneficial to a specific target domain, either from a single or multiple pretrained GANs. This is done using a miner network that identifies which part of the generative distribution of each pretrained GAN outputs samples closest to the target domain. Mining effectively steers GAN sampling towards suitable regions of the latent space, which facilitates the posterior finetuning and avoids pathologies of other methods, such as mode collapse and lack of flexibility. Furthermore, to prevent overfitting on small target domains, we introduce sparse subnetwork selection, that restricts the set of trainable neurons to those that are relevant for the target dataset. We perform comprehensive experiments on several challenging datasets using various GAN architectures (BigGAN, Progressive GAN, and StyleGAN) and show that the proposed method, called MineGAN, effectively transfers knowledge to domains with few target images, outperforming existing methods. In addition, MineGAN can successfully transfer knowledge from multiple pretrained GANs.

READ FULL TEXT

page 7

page 8

page 9

page 10

page 11

page 12

page 14

research
12/11/2019

MineGAN: effective knowledge transfer from GANs to target domains with few images

One of the attractive characteristics of deep neural networks is their a...
research
05/04/2018

Transferring GANs: generating images from limited data

Transferring the knowledge of pretrained networks to new domains by mean...
research
03/10/2023

Feature Unlearning for Generative Models via Implicit Feedback

We tackle the problem of feature unlearning from a pretrained image gene...
research
06/07/2021

GAN Cocktail: mixing GANs without dataset access

Today's generative models are capable of synthesizing high-fidelity imag...
research
02/26/2020

On Leveraging Pretrained GANs for Limited-Data Generation

Recent work has shown GANs can generate highly realistic images that are...
research
02/18/2018

RadialGAN: Leveraging multiple datasets to improve target-specific predictive models using Generative Adversarial Networks

Training complex machine learning models for prediction often requires a...
research
12/16/2021

Ensembling Off-the-shelf Models for GAN Training

The advent of large-scale training has produced a cornucopia of powerful...

Please sign up or login with your details

Forgot password? Click here to reset