AI Paper Atlas

Research map / Image generation / Editing and style transfer

Contrastive Learning for Unpaired Image-to-Image Translation

Taesung Park Alexei A. Efros Richard Zhang Jun-Yan Zhu

ECCV 2020 · Image generation › Editing and style transfer · approach: Image-to-image translation · 1,682 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add weight sharing between encoder and MLP for input and output patches (ablation)CUT · Horse->ZebraFID: 50.5 -> 46.6Improved
add PatchNCE gradient to the decoder (ablation)CUT · Horse->ZebraFID: 444.2 -> 46.6Improved
use CUT instead of CycleGANCUT · CityscapesFID: 76.3 -> 56.4Improved
use CUT instead of CycleGANCUT · Horse->ZebraFID: 77.2 -> 45.5Improved
use CUT instead of CycleGANCUT · Cat->DogFID: 85.9 -> 76.2Improved
use CUT instead of CycleGANCUT · CityscapesmAP: 20.4 -> 24.7Improved
use FastCUT instead of CycleGANFastCUT · Horse->ZebraFID: 77.2 -> 73.4Improved
use FastCUT instead of CycleGANFastCUT · Cat->DogFID: 85.9 -> 94.0Hurt
use CUT (one-sided) instead of CycleGAN, training timeCUT · Horse->Zebratraining speed: 0.40 -> 0.24 sec/iterImproved
add identity loss PatchNCE on domain Y (ablation)CUT · Horse->Zebra / CityscapesFIDMixed
sample negatives from within the input image instead of external or added external negatives (ablation)CUT · Cityscapes / Horse->ZebraFIDImproved
use multiple encoder layers instead of last layer only (ablation)CUT · Cityscapes / Horse->ZebraFIDImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy