Research map / Image generation / Autoregressive and latent generation
SpotEdit: Selective Region Editing in Diffusion Transformers
Zhibin Qin, Zhenxiong Tan, Zeqing Wang, Songhua Liu, Xinchao Wang
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 periodic token reset (ablation) | SpotEdit · PIE-Bench++ | PSNR: 17.10 -> 18.73 | Improved |
| add periodic token reset (speed cost) (ablation) | SpotEdit · PIE-Bench++ | speedup: 2.25 -> 1.95 | Hurt |
| cache condition image features instead of recomputing (ablation) | SpotEdit · PIE-Bench++ | speedup: 1.24 -> 1.95 | Improved |
| cache condition image features instead of recomputing (ablation) | SpotEdit · PIE-Bench++ | PSNR: 19.155 -> 18.73 | Hurt |
| use SpotFusion instead of naive skip or static token fusion (ablation) | SpotEdit · qualitative (Figure 6) | artifacts | Improved |
| apply SpotEdit to FLUX.1-Kontext | SpotEdit · imgEdit-Benchmark | speedup: 1.00x -> 1.67x | Improved |
| apply SpotEdit to FLUX.1-Kontext | SpotEdit · PIE-Bench++ | PSNR: 18.76 -> 18.73 | No clear effect |
| SpotEdit versus TeaCache/TaylorSeer on speed | SpotEdit · imgEdit-Benchmark | speedup: 3.43x (TeaCache) -> 1.67x | Hurt |
| apply SpotEdit to Qwen-Image-Edit | SpotEdit · PIE-Bench++ (Qwen) | PSNR: 18.32 -> 19.40 | Improved |
| apply SpotEdit to Qwen-Image-Edit | SpotEdit · imgEdit-Benchmark (Qwen) | PSNR: 14.23 -> 14.24 | No clear effect |
| SpotEdit versus FLUX.1-Kontext on VL score | SpotEdit · imgEdit-Benchmark | VL score: 3.91 -> 3.77 | Hurt |
Closest papers by meaning
- DiT4Edit: Diffusion Transformer for Image Editing (AAAI 2025 · similarity 0.90)
- Stable Flow: Vital Layers for Training-Free Image Editing (CVPR 2025 · similarity 0.88)
- High-Fidelity Diffusion-Based Image Editing (AAAI 2024 · similarity 0.87)
- Towards Efficient Diffusion-Based Image Editing with Instant Attention Masks (AAAI 2024 · similarity 0.87)
- Enabling Instructional Image Editing with In-Context Generation in Large Scale Diffusion Transformer (NeurIPS 2025 · similarity 0.87)
- DesignEdit: Unify Spatial-Aware Image Editing via Training-free Inpainting with a Multi-Layered Latent Diffusion Framework (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.
