Decoupled Learning for Conditional Adversarial Networks

01/21/2018
by   Zhifei Zhang, et al.
0

Incorporating encoding-decoding nets with adversarial nets has been widely adopted in image generation tasks. We observe that the state-of-the-art achievements were obtained by carefully balancing the reconstruction loss and adversarial loss, and such balance shifts with different network structures, datasets, and training strategies. Empirical studies have demonstrated that an inappropriate weight between the two losses may cause instability, and it is tricky to search for the optimal setting, especially when lacking prior knowledge on the data and network. This paper gives the first attempt to relax the need of manual balancing by proposing the concept of decoupled learning, where a novel network structure is designed that explicitly disentangles the backpropagation paths of the two losses. Experimental results demonstrate the effectiveness, robustness, and generality of the proposed method. The other contribution of the paper is the design of a new evaluation metric to measure the image quality of generative models. We propose the so-called normalized relative discriminative score (NRDS), which introduces the idea of relative comparison, rather than providing absolute estimates like existing metrics.

READ FULL TEXT

page 2

page 4

page 6

page 7

page 8

research
11/15/2020

CcGAN: Continuous Conditional Generative Adversarial Networks for Image Generation

This work proposes the continuous conditional generative adversarial net...
research
05/24/2018

Cross Domain Image Generation through Latent Space Exploration with Adversarial Loss

Conditional domain generation is a good way to interactively control sam...
research
03/06/2017

Activation Maximization Generative Adversarial Nets

Class label information has been empirically proven to be very useful in...
research
04/26/2020

Evaluation Metrics for Conditional Image Generation

We present two new metrics for evaluating generative models in the class...
research
01/31/2023

Mind the (optimality) Gap: A Gap-Aware Learning Rate Scheduler for Adversarial Nets

Adversarial nets have proved to be powerful in various domains including...
research
07/25/2019

Y-Autoencoders: disentangling latent representations via sequential-encoding

In the last few years there have been important advancements in generati...
research
04/03/2018

DeSIGN: Design Inspiration from Generative Networks

Can an algorithm create original and compelling fashion designs to serve...

Please sign up or login with your details

Forgot password? Click here to reset