Research map / Image generation / Text-to-image diffusion
Visual-Relation Conscious Image Generation from Structured-Text
Duc Minh Vo, Akihiro Sugimoto
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 the individual usage subnet (ablation) | visual-relation layout module · COCO-stuff (256x256) | FID: 55.21 -> 27.39 | Improved |
| add weighted unification instead of averaging BBs (ablation) | visual-relation layout module · COCO-stuff (256x256) | FID: 49.16 -> 27.39 | Improved |
| add refinedBB2layout when building the layout (ablation) | visual-relation layout module · COCO-stuff (256x256) | FID: 46.55 -> 27.39 | Improved |
| add pixel loss L_pix (ablation) | full model · COCO-stuff (256x256) | FID: 43.18 -> 27.39 | Improved |
| add context loss L_context (ablation) | full model · COCO-stuff (256x256) | FID: 40.10 -> 27.39 | Improved |
| add adversarial loss L_adv (ablation) | full model · COCO-stuff (256x256) | FID: 41.05 -> 27.39 | Improved |
| add the individual usage subnet (ablation) | visual-relation layout module · COCO-stuff | rIoU: 0.2652 -> 0.3186 | Improved |
| add weighted unification instead of averaging BBs (ablation) | visual-relation layout module · COCO-stuff | rIoU: 0.2795 -> 0.3186 | Improved |
| add the individual usage subnet (ablation) | visual-relation layout module · GENOME (256x256) | FID: 71.49 -> 36.79 | Improved |
| add the individual usage subnet (ablation) | visual-relation layout module · GENOME | rIoU: 0.1264 -> 0.1847 | Improved |
| stack GANs up to 256x256 instead of stopping at 64x64 (ablation) | stacking-GANs · COCO-stuff | IS: 9.20 -> 14.20 | Improved |
| stack GANs up to 256x256 instead of stopping at 64x64 (ablation) | stacking-GANs · COCO-stuff | FID: 35.12 -> 27.39 | Improved |
Closest papers by meaning
- Exploiting Relationship for Complex-scene Image Generation (AAAI 2021 · similarity 0.89)
- Learning to Compose Visual Relations (NeurIPS 2021 · similarity 0.87)
- StackGAN++: Realistic Image Synthesis with Stacked Generative Adversarial Networks (TPAMI 2018 · similarity 0.86)
- StackGAN: Text to Photo-Realistic Image Synthesis With Stacked Generative Adversarial Networks (ICCV 2017 · similarity 0.86)
- Relation-aware Hierarchical Prompt for Open-vocabulary Scene Graph Generation (AAAI 2025 · similarity 0.85)
- HDLayout: Hierarchical and Directional Layout Planning for Arbitrary Shaped Visual Text Generation (AAAI 2025 · 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.
