Unsupervised Image-To-Image Translation

69 papers with code • 2 benchmarks • 2 datasets

Unsupervised image-to-image translation is the task of doing image-to-image translation without ground truth image-to-image pairings.

( Image credit: Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks )

Libraries

Use these libraries to find Unsupervised Image-To-Image Translation models and implementations

Most implemented papers

Improving Shape Deformation in Unsupervised Image-to-Image Translation

brownvc/ganimorph ECCV 2018

Unsupervised image-to-image translation techniques are able to map local texture between two domains, but they are typically unsuccessful when the domains require larger shape change.

CyCADA: Cycle-Consistent Adversarial Domain Adaptation

jhoffman/cycada_release ICML 2018

Domain adaptation is critical for success in new, unseen environments.

Dual Contrastive Learning for Unsupervised Image-to-Image Translation

JunlinHan/DCLGAN 15 Apr 2021

Unsupervised image-to-image translation tasks aim to find a mapping between a source domain X and a target domain Y from unpaired training data.

Unsupervised Attention-guided Image to Image Translation

AlamiMejjati/Unsupervised-Attention-guided-Image-to-Image-Translation 6 Jun 2018

Current unsupervised image-to-image translation techniques struggle to focus their attention on individual objects without altering the background or the way multiple objects interact within a scene.

One-Shot Unsupervised Cross Domain Translation

sagiebenaim/OneShotTranslation NeurIPS 2018

Given a single image x from domain A and a set of images from domain B, our task is to generate the analogous of x in B.

Demystifying Inter-Class Disentanglement

avivga/lord ICLR 2020

Learning to disentangle the hidden factors of variations within a set of observations is a key task for artificial intelligence.

Reusing Discriminators for Encoding: Towards Unsupervised Image-to-Image Translation

alpc91/NICE-GAN-pytorch CVPR 2020

The proposed architecture, termed as NICE-GAN, exhibits two advantageous patterns over previous approaches: First, it is more compact since no independent encoding component is required; Second, this plug-in encoder is directly trained by the adversary loss, making it more informative and trained more effectively if a multi-scale discriminator is applied.

Lifespan Age Transformation Synthesis

royorel/Lifespan_Age_Transformation_Synthesis ECCV 2020

Most existing aging methods are limited to changing the texture, overlooking transformations in head shape that occur during the human aging and growth process.

The Surprising Effectiveness of Linear Unsupervised Image-to-Image Translation

eitanrich/lin-im2im 24 Jul 2020

Unsupervised image-to-image translation is an inherently ill-posed problem.

Learning to generate line drawings that convey geometry and semantics

carolineec/informative-drawings CVPR 2022

We introduce a geometry loss which predicts depth information from the image features of a line drawing, and a semantic loss which matches the CLIP features of a line drawing with its corresponding photograph.