Research map / Image generation / Text-to-image diffusion
SCORE: Semantic Collage by Optimizing Rendered Elements
Zefan Shao, Jin Zhou, Hongliang Yang, Pengfei Xu
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 L_geometry to SCORE (ablation) | SCORE · Collage (32 elements) | Aesthetic score: 5.3421 -> 5.6324 | Improved |
| add L_geometry to SCORE (ablation) | SCORE · Collage (32 elements) | CLIP-Contour: 26.0668 -> 27.1836 | Improved |
| add L_geometry to SCORE (ablation) | SCORE · Collage (32 elements) | Overlap ratio: 0.0564 -> 0.0797 | Hurt |
| add L_overlap to SCORE (ablation) | SCORE · Collage (32 elements) | Overlap ratio: 0.4088 -> 0.0797 | Improved |
| add L_overlap to SCORE (ablation) | SCORE · Collage (32 elements) | CLIP-RGB / CLIP-Contour: 23.2130 -> 22.1625; 27.7504 -> 27.1836 | Hurt |
| add L_overlap to SCORE (ablation) | SCORE · Collage (32 elements) | Aesthetic score: 5.3640 -> 5.6324 | Improved |
| increase input elements from 16 to 64 | SCORE · Collage | CLIP-Contour: 26.4064 -> 27.5407 | Improved |
| replace CLIP-CLOP with SCORE | text-driven collage · Collage (16 elements) | GPT-4o/VLM score: 5.3 -> 8.2 | Improved |
| replace Image-Space Collage with SCORE | collage · Collage (64 elements) | CLIP-Contour: 25.0503 -> 27.5407 | Improved |
| SCORE vs Image-Space Collage in user study | collage · User study | selection rate (%): 20.8 -> 59 | Improved |
Closest papers by meaning
- CollageNoter: Real-Time and Adaptive Collage Layout Design for Screenshot-Based E-Note-Taking (AAAI 2025 · similarity 0.86)
- OneIG-Bench: Omni-dimensional Nuanced Evaluation for Image Generation (NeurIPS 2025 · similarity 0.85)
- DesignDiffusion: High-Quality Text-to-Design Image Generation with Diffusion Models (CVPR 2025 · similarity 0.85)
- Alchemist: Turning Public Text-to-Image Data into Generative Gold (NeurIPS 2025 · similarity 0.84)
- Taming Mode Collapse in Score Distillation for Text-to-3D Generation (CVPR 2024 · similarity 0.84)
- Concept Lancet: Image Editing with Compositional Representation Transplant (CVPR 2025 · similarity 0.84)
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.
