Research map / Image generation / Autoregressive and latent generation
Unifying Layout Generation With a Decoupled Diffusion Model
Mude Hui, Zhizheng Zhang, Xiaoyi Zhang, Wenxuan Xie, Yuwang Wang, Yan Lu
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 |
|---|---|---|---|
| Use LDGM instead of UniLayout | Unconditional layout generation · Magazine U-Gen | FID: 36.61 -> 32.73 | Improved |
| Use LDGM instead of UniLayout | Unconditional layout generation · Rico U-Gen | FID: 26.68 -> 26.06 | Improved |
| Use LDGM instead of UniLayout | Unconditional layout generation · PubLayNet U-Gen | FID: 32.29 -> 25.94 | Improved |
| Add parallel, attribute-specific corruption timelines rather than a shared timeline (ablation) | LDGM · Rico Gen-PCM | FID: 29.24 -> 21.59 | Improved |
| Add parallel corruption timelines rather than partially overlapping timelines (ablation) | LDGM · Rico Gen-PCM | FID: 27.71 -> 21.59 | Improved |
| Add parallel corruption timelines rather than sequential timelines (ablation) | LDGM · Rico Gen-PCM | FID: 26.69 -> 21.59 | Improved |
| Use Gaussian rather than uniform noise for position and size (ablation) | LDGM · Rico Gen-PCM | MaxIoU: 0.16 -> 0.59 | Improved |
| Use Gaussian rather than band-diagonal noise for position and size (ablation) | LDGM · Rico Gen-PCM | MaxIoU: 0.29 -> 0.59 | Improved |
| Use confidence-based decoding rather than non-autoregressive decoding (ablation) | LDGM inference · Rico Gen-PCM | FID: 25.14 -> 21.59 | Improved |
| Use confidence-based decoding rather than autoregressive decoding (ablation) | LDGM inference · Rico Gen-PCM | FID: 23.16 -> 21.59 | Improved |
| Add condition flags to identify precise versus corrupted attributes (ablation) | LDGM · Rico Gen-PCM | Retent.: 11.25 -> 99.66 | Improved |
| Use attribute-type-level timelines rather than element-level timelines (ablation) | LDGM · Rico Gen-PCM | FID: 27.71 -> 21.59 | Improved |
Closest papers by meaning
- Towards Aligned Layout Generation via Diffusion Model with Aesthetic Constraints (ICLR 2024 · similarity 0.92)
- Visual Layout Composer: Image-Vector Dual Diffusion Model for Design Layout Generation (CVPR 2024 · similarity 0.92)
- DesignDiffusion: High-Quality Text-to-Design Image Generation with Diffusion Models (CVPR 2025 · similarity 0.88)
- Rethinking The Training And Evaluation of Rich-Context Layout-to-Image Generation (NeurIPS 2024 · similarity 0.87)
- AesthetiQ: Enhancing Graphic Layout Design via Aesthetic-Aware Preference Alignment of Multi-modal Large Language Models (CVPR 2025 · similarity 0.87)
- HiCo: Hierarchical Controllable Diffusion Model for Layout-to-image Generation (NeurIPS 2024 · similarity 0.87)
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.
