Research map / Image generation / GANs
T-GD: Transferable GAN-generated Images Detection Framework
Hyeonseong Jeon, Young Oh Bang, Junyaup Kim, Simon Woo
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 self-training with teacher-controlled regularization (ablation) | T-GD, EfficientNet-B0 · StarGAN -> PGGAN | AUROC (%): 92.54% to 94.94% | Improved |
| add self-training with teacher-controlled regularization (ablation) | T-GD, EfficientNet-B0 · StarGAN source performance after transfer to PGGAN | AUROC (%): 98.96% to 99.15% | Improved |
| add intra-class Cutmix, JPEG compression, Gaussian blur, and random horizontal flip (ablation) | T-GD, EfficientNet-B0 · PGGAN -> StyleGAN2 | AUROC (%): 99.38% to 98.13% | Hurt |
| add intra-class Cutmix, JPEG compression, Gaussian blur, and random horizontal flip (ablation) | T-GD, EfficientNet-B0 · PGGAN source performance after transfer to StyleGAN2 | AUROC (%): 85.04% to 95.08% | Improved |
| replace inter-class Cutmix with intra-class Cutmix (ablation) | T-GD · StarGAN -> PGGAN | validation loss | Improved |
| use EfficientNet-B0 instead of ResNext32×4d as the base model | T-GD · GAN-image transfer tasks | AUROC (%) | Mixed |
Closest papers by meaning
- Towards Discovery and Attribution of Open-World GAN Generated Images (ICCV 2021 · similarity 0.89)
- Learning on Gradients: Generalized Artifacts Representation for GAN-Generated Images Detection (CVPR 2023 · similarity 0.88)
- GLeaD: Improving GANs With a Generator-Leading Task (CVPR 2023 · similarity 0.88)
- TransGAN: Two Pure Transformers Can Make One Strong GAN, and That Can Scale Up (NeurIPS 2021 · similarity 0.87)
- E$^2$GAN: Efficient Training of Efficient GANs for Image-to-Image Translation (ICML 2024 · similarity 0.86)
- RG-GAN: Dynamic Regenerative Pruning for Data-Efficient Generative Adversarial Networks (AAAI 2024 · similarity 0.85)
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.
