Early Timestep Zero-Shot Candidate Selection for Instruction-Guided Image Editing
Joowon Kim, Ziseok Lee, Donghyeon Cho, Sanghyun Jo, Yeonsung Jung Kyungsu Kim Eunho Yang · 2025
Read Early Timestep Zero-Shot Candidate Selection for Instruction-Guided Image Editing by Joowon Kim, Ziseok Lee, Donghyeon Cho, Sanghyun Jo, Yeonsung Jung Kyungsu Kim Eunho Yang (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-13490.txt.
Description
Joowon Kim and co-authors present ELECT in 2025, a way to pick good seeds early for instruction guided image editing. Diffusion editors often spoil the background while making the asked change, and users retry seeds or prompts by trial and error. Checking many full edits is slow, and outside verifier tools are not always on hand.
ELECT scores seeds at the first diffusion steps by how much the background already drifts. Seeds that keep the background while staying editable rise to the top, and poor ones are cut early. A baseline built on background scores shows the best of many seeds can be strong without supervision. ELECT also plugs into editing pipelines and pairs with large multimodal models to pick seed and prompt together, aiming for fewer wasted full runs.
Catalog record
| CREATOR | Joowon Kim, Ziseok Lee, Donghyeon Cho, Sanghyun Jo, Yeonsung Jung Kyungsu Kim Eunho Yang |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 57 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-13490