Research map / Image generation / GANs
Structured Generative Adversarial Networks
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 SGAN instead of TripleGAN | semi-supervised classifier · MNIST test, n=20 labels | classification error (%): 5.40($\pm$6.53) -> 4.0($\pm$4.14) | Helped |
| Use SGAN instead of TripleGAN | semi-supervised classifier · MNIST test, n=50 labels | classification error (%): 1.59($\pm$0.69) -> 1.29($\pm$0.47) | Helped |
| Use SGAN instead of Ladder | semi-supervised classifier · MNIST test, n=100 labels | classification error (%): 0.89($\pm$0.50) -> 0.89($\pm$0.11) | No clear effect |
| Use SGAN instead of TripleGAN | semi-supervised classifier · SVHN test, n=1000 labels | classification error (%): 5.83($\pm$0.20) -> 5.73($\pm$0.12) | Helped |
| Use SGAN instead of ALI | semi-supervised classifier · CIFAR-10 test, n=4000 labels | classification error (%): 18.3 -> 17.26($\pm$0.69) | Helped |
| Use SGAN instead of TripleGAN | conditional generator · MNIST generated samples, n=20 labels | classification error (%): 3.06 -> 1.68 | Helped |
| Use SGAN instead of TripleGAN | conditional generator · MNIST generated samples, n=50 labels | classification error (%): 1.80 -> 1.23 | Helped |
| Use SGAN instead of TripleGAN | conditional generator · MNIST generated samples, n=100 labels | classification error (%): 1.29 -> 0.93 | Helped |
| Use SGAN instead of TripleGAN | conditional generator · CIFAR-10 generated images | Inception Score: 5.08($\pm$0.09) -> 6.91($\pm$0.07) | Helped |
| Add R_y to SGAN (ablation) | SGAN without collaborative games · MNIST generation | controllability | Helped |
| Add R_z to SGAN (ablation) | SGAN without R_z · MNIST generation | controllability | Helped |
| Add R_y to SGAN (ablation) | SGAN without R_y · MNIST generation | controllability | Helped |
Closest papers by meaning
- TR0N: Translator Networks for 0-Shot Plug-and-Play Conditional Generation (ICML 2023 · similarity 0.91)
- Conditional Generative Adversarial Network for Structured Domain Adaptation (CVPR 2018 · similarity 0.90)
- Taming the Tail in Class-Conditional GANs: Knowledge Sharing via Unconditional Training at Lower Resolutions (CVPR 2024 · similarity 0.90)
- Diversity-Sensitive Conditional Generative Adversarial Networks (ICLR 2019 · similarity 0.90)
- Improved Techniques for Training GANs (NeurIPS 2016 · similarity 0.90)
- Attack Deterministic Conditional Image Generative Models for Diverse and Controllable Generation (AAAI 2024 · 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.