Research map / Image generation / Text-to-image diffusion
AnyCanvas: Potential Field Guidance for Training-Free Spatial Control in Text-to-Image Diffusion
Tianyi Xie, Zhiyuan Yu, kaihong huang, Beilun Wang, Zhaoyang Wang, Dian Shen
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 translation module (ablation) | AnyCanvas on SDXL · PTP | ICQS: 31.38 -> 32.68 | Improved |
| add scaling module (ablation) | AnyCanvas on SDXL · PTP | ICQS: 31.84 -> 32.68 | Improved |
| add translation module (ablation) | AnyCanvas on SDXL · PTP | Canvas-IoU: 35.09 -> 36.44 | Improved |
| add scaling module (ablation) | AnyCanvas on SDXL · PTP | Canvas-IoU: 35.78 -> 36.44 | Improved |
| add AnyCanvas guidance to SDXL | SDXL · PTP | Canvas-IoU: 34.48 -> 36.44 | Improved |
| add AnyCanvas guidance to SDXL | SDXL · PTP | CLIP score: 32.75 -> 32.55 | No clear effect |
| add AnyCanvas guidance to SD2.0 | SD2.0 · PTP | ICQS: 24.87 -> 26.04 | Improved |
| add AnyCanvas guidance to SD1.4 | SD1.4 · PTP | ICQS: 26.74 -> 26.88 | No clear effect |
| add AnyCanvas guidance to PixArt-alpha DiT | PixArt-alpha · PTP | ICQS: 31.39 -> 32.69 | Improved |
| use Flux.1 with linguistic mask prompts instead of SDXL | Flux.1 · DiffusionDB | CLIP score: 32.39 -> 29.28 | Hurt |
| add AnyCanvas to SDXL for content suppression | SDXL · Content suppression | Canvas-IoU: 20.52 -> 22.50 | Improved |
| add AnyCanvas to SDXL for content suppression | SDXL · Content suppression | ICQS: 24.15 -> 25.21 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- TextCenGen — TextCenGen: Attention-Guided Text-Centric Background Adaptation for Text-to-Image Generation (ICML 2025) · better on 6 of 6 shared results
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
- Universal Few-shot Spatial Control for Diffusion Models (NeurIPS 2025 · similarity 0.87)
- Any-Size-Diffusion: Toward Efficient Text-Driven Synthesis for Any-Size HD Images (AAAI 2024 · similarity 0.87)
- R&B: Region and Boundary Aware Zero-shot Grounded Text-to-image Generation (ICLR 2024 · similarity 0.86)
- FreeScene: Mixed Graph Diffusion for 3D Scene Synthesis from Free Prompts (CVPR 2025 · similarity 0.86)
- CtrLoRA: An Extensible and Efficient Framework for Controllable Image Generation (ICLR 2025 · similarity 0.86)
- SSMG: Spatial-Semantic Map Guided Diffusion Model for Free-Form Layout-to-Image Generation (AAAI 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.
