Research map / Image generation / Autoregressive and latent generation
Shadow Generation for Composite Image in Real-World Scenes
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 SGRNet instead of the strongest listed baseline | DESOBA BOS test images · DESOBA BOS test images | LRMSE: 75.099 -> 61.763 | Helped |
| use SGRNet instead of the strongest listed baseline | DESOBA BOS-free test images · DESOBA BOS-free test images | LRMSE: 81.444 -> 61.439 | Helped |
| add a separate background encoder (ablation) | SGRNet · DESOBA BOS test images | LRMSE: 68.876 -> 61.763 | Helped |
| add CAI for foreground-background feature interaction (ablation) | SGRNet · DESOBA BOS test images | LRMSE: 68.031 -> 61.763 | Helped |
| provide the background object-shadow mask (ablation) | SGRNet · DESOBA BOS test images | LRMSE: 63.141 -> 61.763 | Helped |
| use the illumination-based shadow filling composition system (ablation) | SGRNet · DESOBA BOS test images | LRMSE: 67.789 -> 61.763 | Helped |
| supervise predicted shadow parameters (ablation) | SGRNet · DESOBA BOS test images | LRMSE: 65.054 -> 61.763 | Helped |
| condition the discriminator on object and shadow masks (ablation) | SGRNet · DESOBA BOS test images | LRMSE: 65.238 -> 61.763 | Helped |
| add generator adversarial loss (ablation) | SGRNet · DESOBA BOS test images | LRMSE: 67.056 -> 61.763 | Helped |
Closest papers by meaning
- Shadow Generation with Decomposed Mask Prediction and Attentive Shadow Filling (AAAI 2024 · similarity 0.93)
- Shadow Generation for Composite Image Using Diffusion Model (CVPR 2024 · similarity 0.93)
- Shadow Generation Using Diffusion Model with Geometry Prior (CVPR 2025 · similarity 0.89)
- MetaShadow: Object-Centered Shadow Detection, Removal, and Synthesis (CVPR 2025 · similarity 0.86)
- Polarization Guided Mask-Free Shadow Removal (AAAI 2025 · similarity 0.84)
- OmniSR: Shadow Removal Under Direct and Indirect Lighting (AAAI 2025 · similarity 0.83)
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.