Research map / Image generation / Text-to-image diffusion
Dual Optimal Transport for Multi-Concept Composition: Structural Alignment and Texture Injection in Diffusion Models
Hao Fu, Tianyu Su, Meng Liu, Chenfang Yang, Tian Gan
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 Barycentric Soft-Transport structure sketching on top of layout planning (ablation) | OTComp · qualitative (Figure 5) | visual alignment of color, shape, pose | Improved |
| add Geometry-Guided Transport texture rendering after structure sketching (ablation) | OTComp · qualitative (Figure 5) | high-frequency identity detail | Improved |
| use barycentric projection guidance instead of directly minimizing Sinkhorn loss (ablation) | Barycentric Soft-Transport · qualitative (Figure 6a) | structural stability | Improved |
| use OT-based transport instead of rigid Spatial MSE feature matching (ablation) | Barycentric Soft-Transport · qualitative (Figure 6b) | identity under pose change | Improved |
| apply structural guidance early t in [5, 20] rather than late t < 10 (ablation) | Barycentric Soft-Transport · qualitative (Figure 6c) | texture preservation | Improved |
| add Barycentric Soft-Transport to MRSA attention (Ours+MRSA vs SDXL+MRSA) (ablation) | OTComp · qualitative (Figure 7) | semantic leakage | Improved |
| replace MRSA with concept-wise masked attention CWSA (ablation) | OTComp · qualitative (Figure 7) | semantic leakage and identity | Improved |
| use OTComp vs training-free and tuned baselines on four-concept composition | OTComp · MagicBench (four concepts) | DINO: 0.5175 -> 0.6844 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- OminiControl — OminiControl: Minimal and Universal Control for Diffusion Transformer (ICCV 2025) · better on 1 of 1 shared result
- FreeCustom — FreeCustom: Tuning-Free Customized Image Generation for Multi-Concept Composition (CVPR 2024) · better on 1 of 1 shared result
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
- One-Shot Reference-based Structure-Aware Image to Sketch Synthesis (AAAI 2025 · similarity 0.88)
- TextureDreamer: Image-Guided Texture Synthesis Through Geometry-Aware Diffusion (CVPR 2024 · similarity 0.86)
- TexTailor: Customized Text-aligned Texturing via Effective Resampling (ICLR 2025 · similarity 0.86)
- Text-Guided 3D Face Synthesis - From Generation to Editing (CVPR 2024 · similarity 0.86)
- Training-free Dense-Aligned Diffusion Guidance for Modular Conditional Image Synthesis (CVPR 2025 · similarity 0.86)
- Single Mesh Diffusion Models with Field Latents for Texture Generation (CVPR 2024 · similarity 0.86)
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.
