Research map / Image generation / Text-to-image diffusion
Setting the Stage: Text-Driven Scene-Consistent Image Generation
Cong Xie, Che Wang, Yan Zhang, Ruiqi Yu, Han Zou, Zheng Pan, Zhenpeng Zhan
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 correspondence-guided attention loss (ablation) | Qwen-Image-Edit scene staging · Scene staging benchmark (425 images, 2125 pairs) | G2.5F-SA: 9.532 -> 9.811 | Improved |
| add correspondence-guided attention loss (ablation) | Qwen-Image-Edit scene staging · Scene staging benchmark (425 images, 2125 pairs) | G2.5F-TIA: 8.569 -> 9.639 | Improved |
| train on our video-based scene data instead of DL3DV-10K (ablation) | Qwen-Image-Edit scene staging · Scene staging benchmark (425 images, 2125 pairs) | G2.5F-SA: 9.048 -> 9.811 | Improved |
| train on our video-based scene data instead of DL3DV-10K (ablation) | Qwen-Image-Edit scene staging · Scene staging benchmark (425 images, 2125 pairs) | G2.5F-TIA: 8.960 -> 9.639 | Improved |
| add MINIMA-based filtering of generated frames (ablation) | Qwen-Image-Edit scene staging · Scene staging benchmark (425 images, 2125 pairs) | G2.5F-SA: 9.648 -> 9.811 | Improved |
| add MINIMA-based filtering of generated frames (ablation) | Qwen-Image-Edit scene staging · Scene staging benchmark (425 images, 2125 pairs) | G2.5F-TIA: 9.079 -> 9.639 | Improved |
| use G2.5F-TIA instead of CLIP-T as text-image metric | metric choice · Metric validation | pairwise accuracy vs humans: 41.64% -> 83.36% | Improved |
| use G2.5F-SA instead of DreamSim as scene metric | metric choice · Metric validation | pairwise accuracy vs humans: 90.85% -> 95.63% | Improved |
Closest papers by meaning
- LLM Blueprint: Enabling Text-to-Image Generation with Complex and Detailed Prompts (ICLR 2024 · similarity 0.90)
- SpotActor: Training-Free Layout-Controlled Consistent Image Generation (AAAI 2025 · similarity 0.89)
- Generative Photography: Scene-Consistent Camera Control for Realistic Text-to-Image Synthesis (CVPR 2025 · similarity 0.89)
- 3D-SceneDreamer: Text-Driven 3D-Consistent Scene Generation (CVPR 2024 · similarity 0.89)
- SceneDecorator: Towards Scene-Oriented Story Generation with Scene Planning and Scene Consistency (NeurIPS 2025 · similarity 0.89)
- SSMG: Spatial-Semantic Map Guided Diffusion Model for Free-Form Layout-to-Image Generation (AAAI 2024 · similarity 0.89)
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.
