Research map / Image generation / GANs
Diversity-Sensitive Conditional 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 |
|---|---|---|---|
| Add diversity regularization to cGAN | cGAN image-to-image translation · label→image | LPIPS: 0.01 -> 0.18 | Helped |
| Add diversity regularization to cGAN | cGAN image-to-image translation · label→image | FID: 85.07 -> 57.20 | Helped |
| Add diversity regularization to cGAN | cGAN image-to-image translation · map→image | LPIPS: 0.02 -> 0.13 | Helped |
| Add diversity regularization to cGAN | cGAN image-to-image translation · map→image | FID: 90.08 -> 49.92 | Helped |
| Add diversity regularization to cGAN | cGAN image-to-image translation · edge→photo | LPIPS: 0.02 -> 0.12 | Helped |
| Add diversity regularization to cGAN | cGAN image-to-image translation · edge→photo | FID: 31.80 -> 23.06 | Mixed |
| Replace output-space distance with discriminator feature-space distance (ablation) | DSGAN inpainting · CelebA image inpainting | LPIPS: 0.01 -> 0.05 | Helped |
| Replace output-space distance with discriminator feature-space distance (ablation) | DSGAN inpainting · CelebA image inpainting | FID: 13.95 -> 13.94 | Helped |
| Increase regularization weight to λ=20 (ablation) | DSGAN label→image · label→image | FID: FID=191 | Hurt |
| Increase regularization weight to λ=20 (ablation) | DSGAN label→image · label→image | LPIPS: LPIPS=0.20 | Helped |
| Add diversity regularization to pix2pixHD | pix2pixHD · Cityscapes | LPIPS: 0.00 -> 0.12 | Helped |
| Add diversity regularization to pix2pixHD | pix2pixHD · Cityscapes | Segmentation accuracy (%): 0.93 -> 0.92 | Hurt |
Closest papers by meaning
- Mode Seeking Generative Adversarial Networks for Diverse Image Synthesis (CVPR 2019 · similarity 0.95)
- Hierarchical Modes Exploring in Generative Adversarial Networks (AAAI 2020 · similarity 0.93)
- Robust Conditional Generative Adversarial Networks (ICLR 2019 · similarity 0.93)
- Attack Deterministic Conditional Image Generative Models for Diverse and Controllable Generation (AAAI 2024 · similarity 0.91)
- CcGAN: Continuous Conditional Generative Adversarial Networks for Image Generation (ICLR 2021 · similarity 0.91)
- Taming the Tail in Class-Conditional GANs: Knowledge Sharing via Unconditional Training at Lower Resolutions (CVPR 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.