Research map / Image generation / Text-to-image diffusion
NEAF: Natural Image Editing with Attention Fusion for Generalizable Test-time Optimization in Text-Guided Image Editing
Jisoo Kim, Heeseok Oh
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 |
|---|---|---|---|
| use NEAF test-time attention fusion vs best tuning-free baseline LEDITs++ | NEAF · TEdBench (VIEScore GPT-4o) | Overall O: 7.26 -> 7.36 | Improved |
| use NEAF vs Imagic | NEAF · TEdBench (VIEScore Gemini) | Overall O: 7.26 -> 8.34 | Improved |
| use NEAF vs LEDITs++ | NEAF · TEdBench (VIEScore GPT-4o) | perceptual quality: 7.71 -> 7.00 | Hurt |
| use NEAF vs Imagic | NEAF · TEdBench (VIEScore GPT-4o) | perceptual quality: 7.00 -> 7.00 | No clear effect |
| use NEAF vs Imagic for non-rigid edits | NEAF · computational cost | tuned parameters: 859.46M -> 0.08M | Improved |
| use NEAF vs Imagic | NEAF · computational cost | GFLOPs: 338.75 -> 10.58 | Improved |
| add XA-Conductor module (ablation) | NEAF · qualitative edits | edit fidelity | Improved |
| use BLIP-2 aligned caption as source prompt with null-text inversion (ablation) | NEAF · qualitative edits | edit quality | Improved |
| add CLIP edit loss term weight lambda (ablation) | NEAF · training objective | edit loss | No clear effect |
| NEAF vs Imagic in user study (2AFC) | NEAF · user study, 600 responses | preference rate: exceeding 70% | Improved |
| NEAF vs LEDITs++ in user study (2AFC) | NEAF · user study, 600 responses | preference rate: exceeding 90% | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- LEDITs++ — LEDITS++: Limitless Image Editing using Text-to-Image Models (CVPR 2024) · better on 4 of 5 shared results
- InstructPix2Pix — InstructPix2Pix: Learning To Follow Image Editing Instructions (CVPR 2023) · better on 4 of 4 shared results
- SDEdit — SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations (ICLR 2022) · better on 4 of 4 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
- Towards Understanding Cross and Self-Attention in Stable Diffusion for Text-Guided Image Editing (CVPR 2024 · similarity 0.89)
- Towards Efficient Diffusion-Based Image Editing with Instant Attention Masks (AAAI 2024 · similarity 0.88)
- EditInfinity: Image Editing with Binary-Quantized Generative Models (NeurIPS 2025 · similarity 0.88)
- Inversion-Free Image Editing with Language-Guided Diffusion Models (CVPR 2024 · similarity 0.88)
- BARET: Balanced Attention Based Real Image Editing Driven by Target-Text Inversion (AAAI 2024 · similarity 0.87)
- Textualize Visual Prompt for Image Editing via Diffusion Bridge (AAAI 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.
