Research map / Image generation / GANs
Modular Generative Adversarial Networks
Bo Zhao, Bo Chang, Zequn Jie, Leonid Sigal
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 mask prediction in transformer module (ablation) | ModularGAN · CelebA hair+smile+gender | classification error (%): 56.08 -> 6.09 | Improved |
| add mask prediction in transformer module (ablation) | ModularGAN · CelebA hair+smile | classification error (%): 30.85 -> 4.03 | Improved |
| add mask prediction in transformer module (ablation) | ModularGAN · CelebA gender (single) | classification error (%): 3.58 -> 2.61 | Improved |
| add mask prediction in transformer module (ablation) | ModularGAN · CelebA hair color (single) | classification error (%): 4.01 -> 3.86 | Improved |
| add cyclic loss (ablation) | ModularGAN · CelebA hair+smile+gender | classification error (%): 52.87 -> 6.09 | Improved |
| add cyclic loss (ablation) | ModularGAN · CelebA hair+smile | classification error (%): 25.34 -> 4.03 | Improved |
| add cyclic loss (ablation) | ModularGAN · CelebA hair color (single) | classification error (%): 3.93 -> 3.86 | No clear effect |
| add cyclic loss (ablation) | ModularGAN · CelebA gender (single) | classification error (%): 2.87 -> 2.61 | Improved |
| fixed instead of random order of transformer modules (ablation) | ModularGAN · CelebA hair+smile+gender | classification error (%): 6.23 -> 6.09 | No clear effect |
| fixed instead of random order of transformer modules (ablation) | ModularGAN · CelebA hair+gender | classification error (%): 5.98 -> 6.51 | Mixed |
| fixed instead of random order of transformer modules (ablation) | ModularGAN · CelebA hair+smile | classification error (%): 4.37 -> 4.03 | No clear effect |
| use modular transformers instead of one single multi-domain generator (StarGAN) | multi-attribute translation · CelebA hair+smile+gender | classification error (%): 7.63 -> 6.09 | Improved |
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.
- StarGAN — StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation (CVPR 2018) · better on 12 of 14 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
- StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation (CVPR 2018 · similarity 0.92)
- Polymorphic-GAN: Generating Aligned Samples Across Multiple Domains With Learned Morph Maps (CVPR 2022 · similarity 0.90)
- TraVeLGAN: Image-To-Image Translation by Transformation Vector Learning (CVPR 2019 · similarity 0.90)
- Exploring Explicit Domain Supervision for Latent Space Disentanglement in Unpaired Image-to-Image Translation (TPAMI 2019 · similarity 0.89)
- Breaking the Cycle - Colleagues Are All You Need (CVPR 2020 · similarity 0.89)
- One-Shot Generative Domain Adaptation (ICCV 2023 · similarity 0.89)
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.
