Research map / Image generation / GANs
OOGAN: Disentangling GAN with One-Hot Sampling and Orthogonal Regularization
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use alternating one-hot and continuous sampling (ablation) | OOGAN · CelebA | Perceptual diversity score: 2.44 ±0.05 -> 2.77 ±0.06 | Helped |
| Add orthogonal regularization to Q (ablation) | OOGAN · CelebA | Perceptual diversity score: 2.65 ±0.05 -> 2.77 ±0.06 | Helped |
| Use the compete-free generator (ablation) | OOGAN · CelebA | Perceptual diversity score: 2.69 ±0.03 -> 2.77 ±0.06 | Helped |
| Add orthogonal regularization to Q (ablation) | OOGAN · CelebA | Cos-simil. in Q: 0.21 ± 0.01 -> 0.09 ± 0.03 | Helped |
| Use OOGAN rather than InfoGAN | GAN disentanglement on dSprites · dSprites | Kim et al. disentanglement score: 0.59 ±0.078 -> 0.81 ±0.077 | Helped |
Closest papers by meaning
- InfoGAN-CR and ModelCentrality: Self-supervised Model Training and Selection for Disentangling GANs (ICML 2020 · similarity 0.90)
- Disentangled Image Generation Through Structured Noise Injection (CVPR 2020 · similarity 0.89)
- IB-GAN: Disentangled Representation Learning with Information Bottleneck Generative Adversarial Networks (AAAI 2021 · similarity 0.89)
- Semi-Supervised StyleGAN for Disentanglement Learning (ICML 2020 · similarity 0.87)
- Demystifying Inter-Class Disentanglement (ICLR 2020 · similarity 0.87)
- Overcoming the Disentanglement vs Reconstruction Trade-off via Jacobian Supervision (ICLR 2019 · similarity 0.87)
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.