Harmonizing Maximum Likelihood with GANs for Multimodal Conditional Generation

02/25/2019
by   Soochan Lee, et al.
14

Recent advances in conditional image generation tasks, such as image-to-image translation and image inpainting, are largely accounted to the success of conditional GAN models, which are often optimized by the joint use of the GAN loss with the reconstruction loss. However, we reveal that this training recipe shared by almost all existing methods causes one critical side effect: lack of diversity in output samples. In order to accomplish both training stability and multimodal output generation, we propose novel training schemes with a new set of losses named moment reconstruction losses that simply replace the reconstruction loss. We show that our approach is applicable to any conditional generation tasks by performing thorough experiments on image-to-image translation, super-resolution and image inpainting using Cityscapes and CelebA dataset. Quantitative evaluations also confirm that our methods achieve a great diversity in outputs while retaining or even improving the visual fidelity of generated samples.

READ FULL TEXT

page 4

page 8

page 19

page 20

page 21

page 22

page 24

page 25

research
07/27/2022

Vector Quantized Image-to-Image Translation

Current image-to-image translation methods formulate the task with condi...
research
04/11/2021

CoPE: Conditional image generation using Polynomial Expansions

Generative modeling has evolved to a notable field of machine learning. ...
research
04/14/2019

Biphasic Learning of GANs for High-Resolution Image-to-Image Translation

Despite that the performance of image-to-image translation has been sign...
research
11/10/2021

Palette: Image-to-Image Diffusion Models

We introduce Palette, a simple and general framework for image-to-image ...
research
10/08/2019

Semi Few-Shot Attribute Translation

Recent studies have shown remarkable success in image-to-image translati...
research
04/05/2020

Feature Quantization Improves GAN Training

The instability in GAN training has been a long-standing problem despite...
research
11/25/2022

CHIMLE: Conditional Hierarchical IMLE for Multimodal Conditional Image Synthesis

A persistent challenge in conditional image synthesis has been to genera...

Please sign up or login with your details

Forgot password? Click here to reset