Research map / Image generation / GANs
CircleGAN: Generative Adversarial Learning across Spherical Circles
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 |
|---|---|---|---|
| replace CircleGAN objective with SphereGAN objective (remove circle learning) (ablation) | CircleGAN (s_mult) · CIFAR10 | FID: 12.9 -> 15.8 | Hurt |
| replace l2-normalization with inverse stereographic projection (ablation) | CircleGAN (s_mult) · CIFAR10 | FID: 12.9 -> 20.4 | Hurt |
| use circle-based discriminator instead of SphereGAN | unconditional GAN · CIFAR10 | FID: 17.1 -> 12.2 | Helped |
| use circle-based discriminator instead of SphereGAN | unconditional GAN · STL10 | IS: 9.55 -> 9.18 | Hurt |
| use circle-based discriminator instead of SphereGAN | unconditional GAN · STL10 | FID: 31.4 -> 27.0 | Helped |
| detach auxiliary classifier (ablation) | CircleGAN (s_real) · CIFAR10 conditional | GAN-test: 96.1 -> 84.5 | Hurt |
| detach auxiliary classifier (ablation) | CircleGAN (s_real) · CIFAR100 conditional | GAN-test: 93.9 -> 58.5 | Hurt |
| detach auxiliary classifier (ablation) | CircleGAN (s_mult) · CIFAR10 conditional | FID: 5.83 -> 12.2 | Hurt |
| detach auxiliary classifier (ablation) | CircleGAN (s_real) · CIFAR100 conditional | FID: 7.43 -> 9.45 | Hurt |
| circle-based conditional discriminator instead of Proj. SNGAN | conditional GAN · TinyImagenet | FID: 42.0 -> 15.5 | Helped |
| circle-based conditional discriminator instead of AMGAN | conditional GAN · CIFAR10 | FID: 7.62 -> 5.72 | Helped |
| decay generator learning rate linearly to 0 over 300K iterations | CircleGAN (s_mult) · ImageNet | FID | Helped |
Closest papers by meaning
- SphericGAN: Semi-Supervised Hyper-Spherical Generative Adversarial Networks for Fine-Grained Image Synthesis (CVPR 2022 · similarity 0.87)
- Sphere Generative Adversarial Network Based on Geometric Moment Matching (CVPR 2019 · similarity 0.87)
- Improving 3D-aware Image Synthesis with A Geometry-aware Discriminator (NeurIPS 2022 · similarity 0.85)
- A Geometric Analysis of Deep Generative Image Models and Its Applications (ICLR 2021 · similarity 0.85)
- Structure-preserving GANs (ICML 2022 · similarity 0.85)
- What You See is What You GAN: Rendering Every Pixel for High-Fidelity Geometry in 3D GANs (CVPR 2024 · similarity 0.83)
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.