Research map / Image generation / Editing and style transfer
Semantic Granularity Navigation in Image Editing
Liangsi Lu, Minzhe Guo, Xuhang Chen, Yang Shi
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 |
|---|---|---|---|
| decouple edit progress from scale (Navi controller) plus gate | FlowEdit (SD3) · PIE-Bench | PSNR: 22.46 -> 27.94 | Improved |
| decouple edit progress from scale plus gate | FlowEdit (SD3) · PIE-Bench | Structure distance: 14.64 -> 10.67 | Improved |
| decouple edit progress from scale (no gate) | FlowAlign (SD3) · PIE-Bench | Structure distance: 6.21 -> 5.40 | Improved |
| add optional feasible-region gate to Navi-FlowEdit (ablation) | Navi-FlowEdit · PIE-Bench | PSNR: 22.54 -> 27.94 | Improved |
| add optional feasible-region gate to Navi-FlowEdit (ablation) | Navi-FlowEdit · PIE-Bench | SSIM: 89.36 -> 93.85 | Improved |
| add optional gate to Navi-FlowEdit (ablation) | Navi-FlowEdit · PIE-Bench | CLIP-Whole: 26.01 -> 26.18 | No clear effect |
| decouple scale allocation from progress (ablation) | SD3, K=28 · Cross-backbone | SSIM / PSNR: SSIM 88.22 -> 93.22, PSNR 22.18 -> 27.81 | Improved |
| decouple scale allocation from progress (ablation) | SD3.5, K=28 · Cross-backbone | SSIM / PSNR: SSIM 85.68 -> 92.32, PSNR 22.01 -> 27.45 | Improved |
| decouple scale allocation from progress (ablation) | FLUX.1 [dev], K=28 · Cross-backbone | SSIM / PSNR: SSIM 82.14 -> 91.75, PSNR 21.81 -> 26.83 | Improved |
| decouple scale allocation from progress (ablation) | SD3, K=28 · Cross-backbone | CLIP-Whole / Edited: 26.01 -> 26.15, 22.55 -> 22.67 | No clear effect |
| decouple scale allocation from progress (ablation) | FLUX.1 [dev], K=28 · Cross-backbone | CLIP-Whole / Edited: 27.02 -> 27.06, 23.35 -> 23.42 | No clear effect |
| apply Navi controller to InfEdit and FlowAlign | InfEdit, FlowAlign · ImgEdit-Bench | Basic and UGE averages | 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.
- Tino-Edit — TiNO-Edit: Timestep and Noise Optimization for Robust Diffusion-Based Image Editing (CVPR 2024) · better on 2 of 2 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
- Meta-CoT: Enhancing Granularity and Generalization in Image Editing (CVPR 2026 · similarity 0.88)
- Drag Your Noise: Interactive Point-based Editing via Diffusion Semantic Propagation (CVPR 2024 · similarity 0.88)
- Edit-R2: Context-Aware Reinforcement Learning for Multi-Turn Image Editing (NeurIPS 2026 · similarity 0.87)
- Improving Editability in Image Generation with Layer-wise Memory (CVPR 2025 · similarity 0.87)
- UnicEdit-10M: A Dataset and Benchmark Breaking the Scale-Quality Barrier via Unified Verification for Reasoning-Enriched Edits (CVPR 2026 · similarity 0.87)
- EditInfinity: Image Editing with Binary-Quantized Generative Models (NeurIPS 2025 · similarity 0.87)
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.
