Research map / Image generation / Autoregressive and latent generation
SoftCollage: A Differentiable Probabilistic Tree Generator for Image Collage
Jiahao Yu, Li Chen, Mingrui Zhang, Mading Li
Check 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 SoftCollage instead of VSM | Tree-based image collage generation · AIC train set | Mr: 1.095 -> 1.086 | Improved |
| Use SoftCollage instead of VSM | Tree-based image collage generation · AIC train set | Mn: 0.335 -> 0.284 | Improved |
| Remove the backbone network (ablation) | SoftCollage · AIC train set | Mn: 0.284 -> 0.379 | Hurt |
| Add the backbone network (ablation) | SoftCollage · AIC train set | Mr: 1.107 -> 1.086 | Improved |
| Remove information embedding from the feature extractor (ablation) | SoftCollage · AIC train set | Mr: 1.086 -> 1.254 | Hurt |
| Remove information embedding from the feature extractor (ablation) | SoftCollage · AIC train set | Mn: 0.284 -> 0.252 | Improved |
| Replace the fusion module with feature averaging (ablation) | SoftCollage · AIC train set | Mr: 1.086 -> 30.721 | Hurt |
| Replace the fusion module with feature averaging (ablation) | SoftCollage · AIC train set | Mn: 0.284 -> 0.212 | Improved |
| Remove scaled dot-product attention (ablation) | SoftCollage · AIC train set | Mr: 1.086 -> 1.503 | Hurt |
| Remove scaled dot-product attention (ablation) | SoftCollage · AIC train set | Mn: 0.284 -> 0.278 | Improved |
| Use SoftCollage instead of VSM | Image collage generation · AIC human evaluation, 5-scale | score: 3.088 -> 3.180 | Improved |
| Use SoftCollage instead of VSM | Image collage generation · AIC information conveying test | F1-Score: 0.745 -> 0.810 | Improved |
Closest papers by meaning
- Image Inpainting via Tractable Steering of Diffusion Models (ICLR 2024 · similarity 0.85)
- TreeMeshGPT: Artistic Mesh Generation with Autoregressive Tree Sequencing (CVPR 2025 · similarity 0.84)
- COCOTree: A Dataset and Benchmark for Open Tree-Structured Visual Decomposition (NeurIPS 2026 · similarity 0.83)
- CollageNoter: Real-Time and Adaptive Collage Layout Design for Screenshot-Based E-Note-Taking (AAAI 2025 · similarity 0.82)
- Hierarchical Gaussian Mixture Model Splatting for Efficient and Part Controllable 3D Generation (CVPR 2025 · similarity 0.82)
- ZoomLDM: Latent Diffusion Model for Multi-scale Image Generation (CVPR 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.
