Research map / Image generation / Editing and style transfer
Unpaired Image-to-Image Translation using Adversarial Consistency Loss
Yihao Zhao, Ruihai Wu, Hao Dong
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add adversarial-consistency loss (ablation) | ACL-GAN · Male-to-female (CelebA) | FID: 23.96 -> 16.63 | Improved |
| add adversarial-consistency loss (ablation) | ACL-GAN · Male-to-female (CelebA) | KID: 0.023 ± 0.0003 -> 0.015 ± 0.0003 | Improved |
| add identity loss (ablation) | ACL-GAN · Male-to-female (CelebA) | FID: 18.39 -> 16.63 | Improved |
| add identity loss (ablation) | ACL-GAN · Male-to-female (CelebA) | KID: 0.018 ± 0.0004 -> 0.015 ± 0.0003 | Improved |
| add bounded focus mask and mask loss (ablation) | ACL-GAN · Male-to-female (CelebA) | FID: 17.72 -> 16.63 | Improved |
| add bounded focus mask and mask loss (ablation) | ACL-GAN · Male-to-female (CelebA) | KID: 0.017 ± 0.0004 -> 0.015 ± 0.0003 | Improved |
| replace CouncilGAN with ACL-GAN | CouncilGAN · Glasses removal (CelebA) | FID: 27.77 -> 23.72 | Improved |
| replace CouncilGAN with ACL-GAN | CouncilGAN · Male-to-female (CelebA) | FID: 18.10 -> 16.63 | Improved |
| replace CouncilGAN with ACL-GAN | CouncilGAN · Selfie-to-anime (selfie2anime) | FID: 98.87 -> 93.58 | Improved |
| replace MUNIT (cycle-consistency) with ACL-GAN | MUNIT · Glasses removal (CelebA) | FID: 28.58 -> 23.72 | Improved |
| replace CycleGAN with ACL-GAN | CycleGAN · Selfie-to-anime (selfie2anime) | FID: 102.92 -> 93.58 | Improved |
| learn translation in both directions synchronously | ACL-GAN · Unpaired translation | performance | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- DRIT++ — DRIT++: Diverse Image-to-Image Translation via Disentangled Representations (IJCV 2020) · better on 6 of 6 shared results
- StarGAN — StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation (CVPR 2018) · better on 2 of 2 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
- Unpaired Image-To-Image Translation Using Cycle-Consistent Adversarial Networks (ICCV 2017 · similarity 0.92)
- Semantically Robust Unpaired Image Translation for Data With Unmatched Semantics Statistics (ICCV 2021 · similarity 0.89)
- Breaking the Cycle - Colleagues Are All You Need (CVPR 2020 · similarity 0.89)
- Unsupervised Image-to-Image Translation with Stacked Cycle-Consistent Adversarial Networks (ECCV 2018 · similarity 0.89)
- Self-Supervised Dense Consistency Regularization for Image-to-Image Translation (CVPR 2022 · similarity 0.89)
- Instance-Wise Hard Negative Example Generation for Contrastive Learning in Unpaired Image-to-Image Translation (ICCV 2021 · similarity 0.88)
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.
