Research map / Image generation / Text-to-image diffusion
Step-by-step Layered Design Generation
Faizan Farooq Khan, Joseph K J, Koustava Goswami, Mohamed Elhoseiny, Balaji Vasan Srinivasan
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 SLEDGE instead of T2I+LLMGA | step-by-step design generation · IDeation Benchmark | theme adherence: 3.45 -> 3.57 | Improved |
| use SLEDGE instead of iCOLE | step-by-step design generation · IDeation Benchmark | edit compliance: 2.01 -> 3.12 | Improved |
| use SLEDGE instead of T2I+LLMGA | step-by-step design generation · Crello | edit compliance: 2.87 -> 3.60 | Improved |
| use SLEDGE instead of cCOLE | final design quality · FID | FID: 86.7 -> 46.8 | Improved |
| use SLEDGE instead of iCOLE | final design quality · text layout | IoU: 4.5 -> 23.5 | Improved |
| use SLEDGE instead of cCOLE | final design quality · text rendering | Text Accuracy: 73.8 -> 89.4 | Improved |
| use SLEDGE instead of baselines in human study | step-by-step design generation · human evaluation, 100 samples | user preference for theme adherence: 63% | Improved |
| fine-tune the alignment module D instead of using the decoder after Step 1 (ablation) | SLEDGE · IDeation Benchmark and Crello | MLLM-judged design quality | Improved |
| apply loss to the full predicted canvas instead of only the MLLM-specified region (ablation) | SLEDGE · IDeation Benchmark and Crello | MLLM-judged design quality | Improved |
| remove optional image input U_t instead of replacing it with an empty canvas (ablation) | SLEDGE · IDeation Benchmark and Crello | MLLM-judged design quality | Improved |
| add the layer extractor module (ablation) | SLEDGE · IDeation Benchmark and Crello | MLLM-judged design quality | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ZONE — ZONE: Zero-Shot Instruction-Guided Local Editing (CVPR 2024) · better on 5 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Towards Aligned Layout Generation via Diffusion Model with Aesthetic Constraints (ICLR 2024 · similarity 0.85)
- Design2GarmentCode: Turning Design Concepts to Tangible Garments Through Program Synthesis (CVPR 2025 · similarity 0.84)
- MaskPLAN: Masked Generative Layout Planning from Partial Input (CVPR 2024 · similarity 0.84)
- Graphic Design with Large Multimodal Model (AAAI 2025 · similarity 0.84)
- Visual Layout Composer: Image-Vector Dual Diffusion Model for Design Layout Generation (CVPR 2024 · similarity 0.82)
- HieraFashDiff: Hierarchical Fashion Design with Multi-stage Diffusion Models (AAAI 2025 · similarity 0.82)
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.
